The last ice age left a distinctive mark on Earth’s higher latitudes. Although it ended 10 000 years ago, landscapes, sediments and the left-over ice caps remain as a treasure trove of data about the evolution of the world and its climate. But there have been other glaciations; the most extensive took place between 600 and 750 million years ago. The signs of ice from so long ago are not immediately apparent; they are locked up in rocks now in tropical deserts such as the Sahara, as well as in places more often associated with ice sheets, such as Greenland and Spitsbergen.
Geologists have built up an enigmatic picture of this great Precambrian ice age; the ancient icy world differs from recent ice ages in both its geography and the processes that created the sedimentary rocks that form the main geological record. Learning how and why this ancient glaciation happened may give us a better understanding of the variations in climate and sea level throughout geological history. With that aim in view, a group of geologists have joined a project of the International Geological Correlation Programme, sponsored by Unesco, with the aim of taking an interdisciplinary approach to the entire glacial record.
Advertisement
Our perception of past glacial processes is coloured by the most recent ice ages . Research concentrates overwhelmingly on the superficial deposits that were spread thinly over the present continents in Quaternary times – the past 1.6 million years. We see mixtures of stones and mud known as till, or less precisely boulder clay, associated with extensive spreads of sand and gravel deposited by glacial streams. Such deposits have little chance of surviving in the geological record – as tillites – for they will often be eroded before they can be buried. The main sequences of sediments affected by the ice accumulate in the seas around ice sheets. Marine glacial rocks survive in the geological record far better than sequences formed on land, although it can be difficult to tell the two types apart. Glacial environments are characterised by interaction of many processes including river and wind action, or currents in the sea. Conclusive evidence for glaciation may be hidden in a complex assemblage of rocks.
Rocks can preserve striking detail of what happened in these ancient glaciations. There are some particularly good examples around Port Askaig, on the Isle of Islay and on the Gurvellachs, a group of islands on the Firth of Lorne in Scotland. Here the rocks are a sequence of sediments laid down by glaciers that entered the sea, interbedded with layers containing wedges of sandstone representing the effects of permafrost on land, and shallow marine sands without a glacial imprint. They record the advance and retreat of ice across a shoreline. An exceptionally thick sequence – 750 metres of rock – shows at least 17 of these cycles.
In a few other places with rocks of the same age, sequences of rocks influenced by glacial action are even thicker – one in Australia is more than 2 kilometres thick. These deposits are far thicker than those left behind by the recent and more familiar ice ages. They give an abundance of evidence of when in the history of the Earth ice was extensive, and how far it extended, across land and sea.
Precambrian rocks more than 570 million years old in China, Norway, Australia and Africa, show traces of several distinct intervals of glaciation. The late Proterozoic glaciation, lasting from about 950 to 570 million years ago, seems to have been the most extensive. All the present-day continents were affected by glaciation, but dating techniques are not precise enough to tell how far the ice spread at any one time. Within the limits of palaeontological and radiometric dating, there seem to have been several discrete glacial periods. In at least one of these, ice seems to have been more extensive than at any time in Earth’s history.
Many late Proterozoic sequences are so well preserved that we can, by careful logging of the strata, establish in considerable detail the environments in which they formed. Furthermore, many of the glacial successions come from icy seas, providing insights into the inaccessible processes of sedimentation around today’s polar ice sheets. This reverses the old geological adage that ‘the present is the key to the past’.
Much of the work on late Proterozoic glacial sediments has taken place in the past 15 years or so, after many years of widespread dismissal of the evidence for this early glaciation. Modern research has provided a detailed picture of the glacial environments in many parts of the world in Proterozoic time, and how such environments are linked to tectonic processes such as mountain building.
To find out what happened in one particular area, several colleagues and I have been unravelling the late Proterozoic evolution of the Arctic-North Atlantic region, especially in east Greenland and Svalbard, a group of islands north of Norway. Thanks to the more recent glaciations in these areas, outcrops are pristine. They provide the clearest possible evidence of ancient glaciation. Despite the challenging conditions in which we have to work, we have confirmed that two main glacial epochs affected both areas during the last 100 million years of Precambrian times. In both areas, the ice left a sequence of rocks formed both on land and in a shallow, icy sea. As a bonus, the deposits form a recognisable sequence that we can find throughout the region, including Scandinavia and Scotland. As neither fossils nor radiometric data give reliable ages for Precambrian rocks in these areas, matching the tillites is a great help in correlating sequences.
Some of the sediments were terrestrial with signs that glaciers had scoured bedrock, others came from coastal areas where glacial sediments intermingle with the ordinary sediments formed in shallow seas and intertidal areas. Still others came from the deep sea, where stones carried away from the shore on rafts of ice dropped into the predominantly sandy and muddy sediments. In other cases, glacial sediments had been reworked in deep water, slumping under their own weight after first being laid down by ice.
Local glaciation, affecting mountainous areas, must be distinguished from general glaciation, if the presence of ice is to be related to changes in the climate. To achieve this, we record accurately the sequence of strata, concentrating on the texture of the rocks (which tells us about the mode of transport), and the character of other, non-glacial, sediments associated with the rocks, especially the carbonates and fossils. Evidence from the Arctic-North Atlantic region suggests a Precambrian ice sheet hundreds of kilometres across. There is no evidence for mountain glaciation; all the rocky debris seems to have been plucked from bedrock beneath glaciers rather than to have tumbled onto the ice from crags above. The ice sheet that we studied in detail was an integral part of the lowland landscape and seashore.
In other parts of the world, tectonic uplift has played a significant role in generating glaciation. As the slow movement of plates on the Earth’s surface produces mountain belts, snow builds up on the new chains, initiating glaciers. Ice flows onto low ground, individual glaciers coalesce into icefields and eventually an ice sheet is born. This is what happened in late Proterozoic times in the US and Canada, in the Western Cordillera and the Appalachian mountains. Glacial sedimentation was sporadic, interrupted by uplift and erosion, as well as by debris flows and volcanic events.
The late Proterozoic glaciations are unique in several ways. Perhaps the most surprising one is that the ice sheets seem to be more important at low latitudes; they were equatorial rather than polar. Evidence for this is supplied by the orientation of the Earth’s magnetic field when the rocks lithified, locked away in grains of magnetic minerals. If the rocks formed close to the equator, the magnetic field lines were roughly horizontal, but if near the poles, the field was much steeper. The surprisingly flat magnetic vectors from Precambrian glacial sediments have been taken to indicate glaciation around the equator, with or without glaciation at the poles. This is a very different situation to that seen in recent ice ages, one that raises the possibility that late Proterozoic climatic zones differed radically from ours.
But this problem may not be quite what it seems; reliable palaeomagnetic data and radiometric ages are scarce, so researchers are not clear exactly where the rocks were and when. These uncertainties have led to widely differing views. One idea, put forward by the Australian scientist George Williams, is that there was more ice at low latitudes because the Earth’s axis was more tilted 600 million years ago. The tropics received less radiation from the Sun than the poles, so ice built up. Another frequently quoted hypothesis holds that the continents were moving especially quickly at this time. There is enough uncertainty in the measurements of latitude and age for ice to have developed at high latitudes as a succession of land masses drifted over the poles, resulting in glacial sediments that are roughly but not exactly the same age over wide areas. Others favour glaciation on every continent irrespective of latitude, with many significant ice sheets or numerous ice caps. Increasingly the evidence seems to support the last hypothesis, although the ice could not have been continuous; according to documented changes in sea level, not enough moisture could be produced on Earth to freeze into such a global ice sheet.
Another enigma of the late Proterozoic glaciations that has puzzled geologists for decades is the close association of tillites with carbonates, mostly limestones and dolomites. Until about 20 years ago, most geologists thought that carbonates formed mainly in warm water. And the mounds and columns of algal limestone known as stromatolites, common in glacial sequences, were taken as further evidence that the carbonates were deposited in warm water. Today the algae that build these structures live mainly in warm water.
But carbonates can form in cold water, as can stromatolites; there are good examples in saline lakes in the Dry Valleys, an area of Antarctica free of ice. Reliance on the warm water hypothesis and failure to recognise cold water carbonates led many geologists to dismiss the many rocks supposed to be tillites as some other type of sediment. For a time in the 1950s and 1960s the entire concept of a glaciation in late Proterozoic time was largely discredited. Geologists explained that the rocks that looked like tillites were the result of tectonic processes, such as earthquakes. For example, they could be debris-flows: masses of rocks and mud that flow down underwater slopes, perhaps triggered by quakes. Those who still clung to some sort of glacial hypothesis held a range of alternative views. These included the existence of warm climates immediately before or after the tillites formed, with the implication of extremely fast changes in climate; strong seasonal variations with warm summers; glaciers eroding older carbonate rocks; glaciation in low latitudes; and, finally, formation of carbonates in the cold waters of a sea or lake. Pre-eminent among those who held fast to the glacial hypothesis was Brian Harland, of the University of Cambridge, whose stand has since been fully vindicated.
Geological opinion has now turned full circle. Detailed sedimentology has indicated that many supposed tillites are indeed of glacial origin, but what of the carbonates? The superbly preserved sequence in the remote highland ice fields of northeastern Spitsbergen hold clues to the answer. Ian Fairchild at the University of Birmingham has joined Baruch Spiro of the British Geological Survey to analyse the petrology and geo-chemistry of both pure carbonates and the carbonate-bearing tillites within the sequence.
Techniques developed to analyse ice and sediments from the most recent ice ages have shown that the proportions of isotopes of oxygen in carbonate can reflect the temperature at which the mineral formed, as well as latitude and the degree of glaciation at the time. At the same time, I have been concentrating on how the glacial sediments were deposited. We have found that the carbonates are not as simple as they may seem; in fact, there are five distinct types. They formed from sea or lake water, with or without microbial action, from the rocks ground up by glaciers, speedy recrystallisation of this ‘rock flour’, and from salty ground water near saline lakes such as those in the Dry Valleys of Antarctica.
Fairchild and Spiro further demonstrated that the proportions of different isotopes of oxygen in these carbonates could provide an indication of the latitude at which the sequences formed. This is based on the premise that present-day rain and snow has increasingly smaller proportions of the heavy isotope of oxygen, oxygen-18, towards the poles. Surprisingly, they found that the Precambrian carbonates were rich enough in the heavy isotope to be inconsistent with polar glaciers. So, as suggested earlier by Harland and Williams, this glaciation probably did reach sea level at a lower latitude than the Quaternary episodes.
The late Proterozoic world must have had a truly ‘icehouse’ climate. We cannot be sure why but it is conceivable that the atmosphere contained less carbon dioxide, reducing the greenhouse effect that warms the planet. And this glaciation, although extensive, was only one of many on the early Earth.
Before about 1000 million years ago, Earth seems to have been more or less free of ice for at least 1000 million years. But between 2500 and 2000 million years ago there was a major phase of glaciation known as the early Proterozoic glacial era. Dating deposits from this long ago is not very precise so we cannot be sure whether glaciation was widespread at any one time, as it appears to have been in the late Proterozoic. The glacial centres may have been relatively small, migrating from place to place as the continents moved across the poles.
There may be even older glacial deposits on Earth, in sediments of the Witwatersrand Supergroup in South Africa. They fall in the latter part of the Archaean Eon, dating from around 2600 million years ago, but add little to the story of the Earth’s ice because they are relatively poorly documented and some geologists doubt their glacial origin.
At present we know little about the causes of these earlier Precambrian glaciations. Tectonic controls, principally the uplift of mountain belts, seem to be important in some cases. Overall, there is no reason to suppose that these early ice ages were any different from recent ones. The few palaeomagnetic data so far obtained suggest that they were relatively high latitude glaciations. It seems that only the late Proterozoic glaciation approached the equator.
Ice ages are not a new phenomenon. Earth’s history is characterised by a series of major glacial phases, when ice sheets covered large areas of land and extended into the sea. Precambrian glaciations were separated by long periods of little or no ice cover, but eventually ice may have covered much of the Earth. Since Precambrian time there have been three further major glaciations around the poles, one between 440 and 420 million years ago and one between 350 and 260 million years ago. The present glacial period began in Antarctica at least 40 million years ago. But one lesson for those concerned about the climate in the future is clear from these early signs of ice: ice on Earth may be as much the norm as more temperate conditions.
Michael Hambrey was at the Scott Polar Research Institute in Cambridge but has recently been appointed professor of Quaternary geology at the Liverpool Polytechnic
* * *
1: READING THE ROCKS FOR SIGNS OF AN ICE AGE
Geologists use a range of evidence to establish whether ice had a hand in the formation of particular strata. Ancient glaciation on land leaves traces in the rocks that match those left by ice in the recent past. Most people living at high latitudes are familiar with the sediments left behind by recent glaciations.
These unconsolidated stony beds have a variety of names. In England, boulder clay is a familiar term, but can be inaccurate as the sediment may contain neither boulders nor clay; even if it does it also has everything else in between, from silt and sand to pebbles and cobbles.
A better term is till, a word of Scottish origin, originally used to refer to stiff, stony ground. But glacial geologists argue about what is a till – for example is it a till if the sediment is deposited from glaciers in water? The answer to this semantic dilemma comes with the term ‘diamicton’, which refers equally to sediments formed by ice on land or sea. The respective terms for the rock are tillite and diamictite.
Glaciations on land have left some distinctive signs. The most convincing evidence for ice sheets and glaciers comes from scraped and abraded rocks. Polished or striated surfaces, scratched by the sand and gravel carried within the glaciers, are as characteristic of glaciations in the distant past as they are of Quaternary or modern ice. Polished and scratched surfaces within sequences of rocks, or on rock fragments, are a sure sign of ice movement. Flowing ice sometimes formed crescent-shaped fractures known as chattermarks or left behind distinctive crescent-shaped gouges.
A second form of evidence comes from sediments left behind as ice moved across the land. Like recent glacial sediments, tillites have very irregular thicknesses, although they are usually less than 50 metres thick, and contain a high proportion of stones and pebbles of all sizes. They often fill in what would otherwise be hollows. Other sedimentary processes can form units of this type, but they usually have distinguishing features. Tillites, for example, may have been deformed by the ice sheet moving above them; this shearing can distort the fine matrix of the rock and may align the stones parallel to the direction in which the ice flowed.
A third line of evidence comes from erosional and depositional landforms that survive in the rock record. The whale-back rocks known as roches moutonnees, perhaps even ancient U-shaped valleys, long filled with sediments and piles of rubble such as moraines and gravel ridges called eskers all indicate past glaciations.
Comparison with glacial features we know today is less straightforward for deposits formed underwater. Sediments in the polar seas laid down in recent glaciations are not easy to reach. The rocks show what types of sediments formed. In a sea or a lake, glacial conditions lead to the formation of poorly sorted beds, containing plenty of stones, with internal forms ranging from structureless to stratified. In stratified beds, dropstone structures are characteristic of glacial conditions. These are what they sound like – structures that result when a stone, carried away from shore on an iceberg, falls into the sea and drops into the sediment. Dropstones disrupt the regular horizontal layers, and their characteristic appearance even when there are no other signs of glaciation is a good indicator of ice in the water.
Individual packages of seafloor sediment may be thicker than those on land, some reaching hundreds of metres, and the boundaries with other rock types may be gradational, rather than sharp as they tend to be on land. The stones they include are normally randomly orientated and the tops of some beds may show the effects of ocean currents, winnowing fine material from the sediment and producing structures that indicate which way was up when the rocks formed.
These sediments often contain fossils; in Precambrian times these were algal mounds, stromatolites, and microscopic plants. Of these, the acritarchs usually indicate formation in the sea, although some may have formed in lakes. Seafloor sediments may also show signs of reworking as debris-flows after deposition, and there may also be finely laminated deposits derived from sediment-laden, high-density currents.
The stones within diamictites are generally a mixture of different rock types; this is the origin of their name. They are not arranged in any particular pattern, by size, for example, or by lithology; large stones lie by small ones, smooth, rounded pebbles are next to sharp angular fragments that look as if they broke off yesterday.
Some of the stones have faceted surfaces and bullet-nose shapes. Often debris has faces showing striations and crescentic fractures, indicating transport at the base of a glacier, where the abrasive effects of the ice are strongest. But debris carried at the surface of the glacier will tend to be predominantly angular. When the ice melts, the worn and fresh fragments mix together to make the characteristic sediments.
On a smaller scale, the minerals in the sediment will be fresh and some, such as garnets and quartz, have distinct fractures when observed under a scanning electron microscope. Some of the stones may be exotic, having no relationship to the rocks below them, nor indeed to any within hundreds of kilometres. These factors point to transport over long distances.
The origin of some stony deposits in some outcrops may not be very clear, but a search for other evidence of a cold climate can help. In particular, geologists need to look for signs of the landforms typical of permafrost today. Frozen ground can shoulder aside stones and soil as it cools and thaws each year, moving stones into distinctive patterns. In rocks, geologists look for wedge-shaped features in cross-section, or, in plan view, for stones sorted into circles, polygons and stripes.
* * *
2: MAKING A CASE FOR ICE AGES
Two hundred years ago, the idea that Europe had once been covered with ice was considered at best a fallacy – and at worst, heresy. Ice ages became respectable scientific fact only through the work of Louis Agassiz in the middle of the 19th century. Observers of the natural world had noticed that huge boulders sometimes lie where they do not belong, far from their native rock formations. They named such stones ‘erratics’. They had also seen the jumbled heaps of rocks and sediment dotted across parts of Europe, but they did not call on ice to explain either.
To most 17th century geologists, it was ‘obvious’ that this debris was the work of the biblical flood, a torrent so devastating that it washed all before it. In 1787, a Swiss clergyman, Bernhard Kuhn, became one of the first to break away from this orthodoxy. He argued that great sheets of ice had moved these erratic boulders.
Among the other scientists who independently arrived at similar ideas was the Scottish geologist James Hutton. When he visited the Alps of France and Switzerland in the 1790s, he was impressed by the scars left on the rocks by glaciers.
Hutton’s espousal of the idea of an ice age was in keeping with his firm belief that the surface of the Earth has been shaped in the distant past by the same processes that we can observe at work today.
This principle of uniformitarianism was revolutionary at the time, for it implied that the Earth had a much longer history than could be inferred from a literal interpretation of the Bible. The very idea created controversy in the late 18th and early 19th centuries. Today, a modified form of uniformitarianism and the existence of the Earth for more than 4.5 billion years are cornerstones of geology.
Uniformitarianism has not survived intact, of course. While accepting that today’s processes operated in the past, geologists now know that their severity and geographical distribution has been very different, and there have been catastrophes every bit as devastating as Noah’s flood. Although it happened on a geological timescale, the spread of ice over much of Europe was one such catastrophe.
It is ironic that Hutton, favouring uniformity rather than catastrophe, was the first to recognise the evidence for such rapid geological change. There is no real conflict with his uniformitarianism, for over a long enough timescale, catastrophe is ‘normal’.
But neither Kuhn nor Hutton, nor anyone else of their generation, made much effort to persuade the scientific world that there really had been an ice age. It remained merely an interesting idea until 1837, when Louis Agassiz, the 30-year-old President of the Swiss Society of Natural Sciences, took up the issue. He promoted it with such vigour, in spite of intense opposition from the biblical catastrophists, that by the middle 1860s the ice age theory had become widely accepted.
The evidence that persuaded Agassiz that ice had scoured the European landscape was all about him in his native Switzerland. But his eyes were opened by the work of Johann von Charpentier, a pioneering glaciologist who was born in 1786 in Freiberg, now part of Germany. Von Charpentier adopted the French version of his name, Jean de Charpentier, after he moved to Switzerland, where he worked as a mining engineer, becoming director of salt mines in Vaud.
De Charpentier set to work with other geologists, mostly amateurs including a civil engineer, Ignatz Venetz-Sitten. In particular, he studied the locations of large boulders from the Swiss Alps that today lie far down the Rhone valley. By 1834, like others before him, de Charpentier had reached the conclusion that these huge, ‘immovable’ boulders could have reached their resting places only in the grip of great glaciers that had, long ago, slid down from the mountains during an ice age. He presented his evidence that year to a meeting of the Society of Natural Sciences, but nobody, including Agassiz, seems to have been convinced.
Indeed, when young Agassiz heard of de Charpentier’s work, he not only didn’t believe a word of it, but set out to prove that the whole idea was nonsense. He was well placed to do so, because in 1832 he had been appointed professor of natural history at the University of Neuchtel, and he was becoming known as an expert on fossil fishes. As a native of Switzerland, he was sure from everyday experience that glaciers could not move far enough, nor fast enough, to transport great lumps of rock down into the Rhone valley.
As a good scientist, Agassiz set out to prove his case. To this end, he established an observing station in a hut on the Aar glacier, and carefully measured the movement of this glacier (among others) by driving stakes into the ice and measuring their movement. Confounded by the discovery that the ice moved much faster than he had thought, and persuaded that it could also carry large boulders, Agassiz was converted from sceptic to believer. Like many converts, he became an enthusiastic evangelist for his new beliefs.
He started, in 1837, by dragging his reluctant fellow members of the Society of Natural Sciences out of the lecture room and into the mountains to see the evidence for themselves. They were not immediately convinced. Some even argued that marks in the rocks might have been made not by the grinding of rocks carried by glaciers, but by the wheels of passing carriages. Agassiz was not to be dissuaded. He went out with Venetz-Sitten and de Charpentier to look at the evidence, but soon raced ahead of them in his enthusiasm, propounding a wide-ranging theory of a world covered by ice, published in 1840.
De Charpentier did not publish his own version until 1841. But even if he had managed to beat Agassiz into print, there is no doubt which version of the story would have made more impact. When Agassiz wrote about ice ages, nobody else could compete. His imagery would give even today’s tabloid journalists a run for their money:
‘The development of these huge ice sheets must have led to the destruction of all organic life at the Earth’s surface. The ground of Europe, previously covered with tropical vegetation and inhabited by herds of giant elephants, enormous hippopotamuses, and gigantic Carnivora became suddenly buried under a vast expanse of ice covering plains, lakes, seas and plateaus alike. The silence of death followed . . . springs dried up, streams ceased to flow, and sunrays rising over that frozen shore . . . were met only by the whistling of northern winds and the rumbling of the crevasses as they opened across the surface of that huge ocean of ice.’
No one, alas, writes scientific papers like that any more; indeed, few scientists did 150 years ago. Although Agassiz did get a little carried away, and life had not completely disappeared from the face of the Earth, let alone from Europe, it was his knack for purple prose and publicity that helped Agassiz to convince his colleagues that ice ages had to be taken seriously. But he would have had no case to make if the pioneers, including de Charpentier, had not painstakingly tracked down and studied the huge erratic boulders scattered around the Alps. Even then, it took the best part of 30 years to convince the doubters.
In 1846, Agassiz visited the United States to study the spectacular remains of glacial activity in the New World. He stayed on as professor of zoology at Harvard University, married Elizabeth Cabot Carey in 1850, and became a pillar of American science until his death in 1873 at the age of 66.
