A new history of Antarctic ice

Antarctic field camp

A field camp for glacial geologists on James Ross Island, northern Antarctic Peninsula.

19 June 2015 by Bethan Davies

There's still a huge amount we don't know about the history of the southern polar ice sheet. Bethan Davies was part of an international team that brought together the latest findings to reveal a complex and dynamic Antarctica.

How sensitive is the Antarctic Ice Sheet to climate change? How does it react to a warmer atmosphere and ocean, and how can we estimate how much ice it is losing now? To answer these questions, we need to know how it has retreated and thinned since the last ice age.

The Scientific Committee for Antarctic Research (SCAR) commissioned an international team of 78 scientists from 14 countries to compile the most complete review of the history of the ice sheet to date. The team includes scientists from many backgrounds, including marine geologists, those studying changes recorded in lake muds, ice-sheet modellers, terrestrial glacial geologists and glaciologists.

The review looked at the ice sheet's extent and thickness at the Last Glacial Maximum - the peak of the last ice age, 25,000 years ago - and then every 5,000 years until the present. The challenge we faced was to take information from many sources and many different international teams, and turn it into a unified history of the Antarctic ice cap over the last few tens of thousands of years.

The Earth is still rebounding after the weight of the ice sheets was removed at the end of the last ice age, and measurements of the modern sheet's height above sea level or changes in mass have to take this into account. This in turn means we need detailed knowledge of past ice volumes. And rates and magnitudes of ice-sheet change during periods of past rapid climate change - such as the last transition between the last ice age and the present interglacial period - and over the last few thousand years put current ice-sheet change into context, and help us understand thresholds and tipping points beyond which even more rapid changes may occur.

The changing ice cap

You might imagine the relationship between ice and temperature is simple - when the climate gets colder the ice increases, and when it warms the ice melts. But our results show the ice sheet is dynamic, with various areas gaining and losing ice at different times.

Around much of the continent, the glaciers reached their peak about 20,000 years ago, when large areas of the ice sheet reached the edge of the continental shelf. Yet in parts of the colder, drier East Antarctic Ice Sheet the story was very different, with the ice sheet remaining relatively stable for the last 25,000 years. In fact, the ice surface in its interior was up to 100m lower than the present day, because the cold, dry air meant there was less snowfall.

In several cases, our work highlighted apparent contradictions that will need more work to understand. For example, the Weddell Sea was a particular area of contention, with apparent discrepancies between the evidence from land and sea. Part of the problem is that this region is remote, even by Antarctic standards, so data are sparse and the extensive sea ice makes investigation from ships difficult.

The limited marine data we do have from the Weddell Sea region suggest that at its maximum, the ice sheet extended to near the outer edge of the continental shelf, implying it was thicker than the present one. In contrast, geological evidence from mountain ranges inshore suggests that there was very little change in the ice sheet's elevation between the Last Glacial Maximum and today. This limited change in thickness indicates that the ice in the Weddell Sea embayment was thin and floating, rather than grounded in the deep troughs on the continental shelf. This leaves scientists with two alternative scenarios to test and investigate over coming years.

Timeline showing the dynamic changes occurring in different parts of the Antarctic continent

Timeline showing the dynamic changes occurring in different parts of the Antarctic continent. It shows measurements of mean annual air temperature compared with the present from the Vostok ice core in East Antarctica. You can clearly see how different parts of the continent reach their maximum ice volume and then deglaciate at different times.

Again and again, we learned that things are more complicated than we'd thought. For example, a key finding is that the ice-sheet's retreat from its maximum extent was asynchronous; that is, it did not respond to a warming climate in a uniform way.

In the west around the Antarctic Peninsula, Bellingshausen Sea and sub-Antarctic islands, the recession was well under way by 18,000 years ago. But at that time most of the East Antarctic Ice Sheet was still grounded near its greatest extent, and the grounding line - the transition between floating ice and ice that rests on the sea floor - in the Ross Sea barely moved at all. By 10,000 years ago, the Antarctic Peninsula Ice Sheet, ice in the Amundsen Sea, and glaciers on the sub-Antarctic islands had largely receded to the inner continental shelf. By 5,000 years ago, their configurations were similar to the present.

Conversely, ice-sheet recession in East Antarctica didn't really get under way until 12,000 to 6,000 years ago, when the oceans were warming significantly. At a more local scale, individual ice streams shrank back at different times, often depending on the landscape they flowed through.

Some of our discoveries have implications far beyond Antarctica. One was that the ice sheet contributed less than 10m to global sea levels as it melted - less than earlier studies estimated - and probably contributed relatively little to the rapid rise of around 20m in global sea levels around 14,500 years ago. Some scientists have thought Antarctica was the source for this massive increase, but our research suggests that other ice sheets probably played a bigger role.

The review reveals a complex Antarctic Ice Sheet that responded sensitively to changes in air and ocean temperature in different ways at different times. Patterns of ice recession varied strongly from place to place, controlled in part by the landscape and ocean water depths. Ice-sheet modellers will now use the project's results to test their computer models on historical data; if the models can successfully simulate what we know happened in the past - known as hindcasting - then we can be more confident in their projections of the future. This will in turn help us understand how the Antarctic ice will respond to a changing environment.

It is clear from the project that the glacial history of large areas of Antarctica remains little-studied and poorly understood, and in some areas we still can't decide between competing hypotheses. It's a complicated story that highlights the complexities involved in reconstructing ancient ice sheets.

Major reviews like this identify the state of the art and clarify future research directions. They give the scientific community access to a wide range of data sources; data that often would not have been available otherwise. Other specialists can then use these data; for example, computer modellers need accurate geological information against which to test their models.

By understanding how, how quickly and how much the ice sheet responded to oceanic and atmospheric changes in the past, we will be better able to judge how it will react to similar changes in the future, and will be able to provide better projections of future sea-level rise.


Bethan Davies is a glaciologist and physical geographer at Royal Holloway, University of London. Email: .

The review described in this article appears as seven papers in a special issue of Quaternary Science Reviews. Funding came from a variety of funding sources, including NERC and SCAR.