For the second time in a month, leading scientists have closely tied the ancient history of the vast Antarctic ice sheet to a key planetary parameter that humans are now controlling - the amount of carbon dioxide in the atmosphere.
Last month, new research showed that during the Miocene era, some 14 to 23 million years ago, Antarctica gave up huge volumes of ice, equivalent to tens of meters of sea level rise, when levels of carbon dioxide in the atmosphere are thought to have been around 500 parts per million. We're at a little over 400 parts per million now.
And now, new research in the journal Science goes back even further, to the ancient era when the ice sheet's growth is believed to have originally begun, some 34 million years ago during a time period known as the Eocene-Oligocene boundary. And it finds that at this time, too, a fall in atmospheric carbon dioxide levels appears to have been involved in allowing glaciation and ice sheet growth.
"Before 33.6 million years ago, there was no ice, and CO2 was above 750, was above the threshold," said Simone Galeotti, lead author of the new study by a large international collaboration of authors, and a researcher at the Università degli Studi di Urbino in Italy. But then came a transitional period with lower carbon dioxide and a variable ice sheet - and then, 32.8 million years ago, carbon dioxide levels dipped below 600 parts per million, and the Antarctic ice sheet greatly expanded.
The research is based on a nearly kilometer-long core of deep ocean sediment taken from the western part of the Ross Sea, not far from the United States' McMurdo station. The distribution of key bits of rock in different layers of the core, said Galeotti, suggests that it wasn't until 32.8 million years ago, emerging from the warmth of the period known as the Eocene, that icebergs were breaking off of an ice sheet that stretched all the way to the Antarctic coast. These icebergs floated out into the sea and melted, and rocks or "clasts" carried in the ice fell to the sea floor, leaving clues that modern geologists can read.
"We see ice-rafted clasts reaching the drilling site only at 32.8 million years. Which says that before that time, the ice sheet was not fully developed," Galeotti said.
Before this period, in the Eocene climate - typically referred to as a "hothouse" world and occurring between 55.8 million and about 34 million years ago - Antarctica actually featured trees and vegetation. The Arctic was also extremely warm back then - alligators swam there.
The planet at the time did not have land-based ice and was far hotter than now, with much higher seas overall and much higher carbon dioxide levels in the atmosphere. This changed in the transition from the Eocene to the Oligocene, which ushered in a cooling that closely tracked declining levels of carbon dioxide in the atmosphere - although it was also buffeted, over long time periods, by tweaks and changes to the Earth's orbit.
"This is one of the biggest climate changes that the planet has seen in the last 50 million years, when Antarctica went from being forested and vegetated, to really . . . it became glaciated at this Eocene-Oligocene boundary," said Rob DeConto, one of the study's co-authors and a geoscientist at the University of Massachusetts, Amherst.
In this context, the new research further tunes our understanding of how much carbon dioxide allows the ice sheet to grow or causes it to melt. "The ice sheet was particularly vulnerable between 33.6 and 32.8 [million years ago], with the CO2 level between 750 and 600," Galeotti said.
That might still sound pretty far off from where we are now, but here's the catch. When the Antarctic ice sheet initially formed, scientists don't believe that it had one of its most distinctive current features - large portions, particularly in West Antarctica but also in key regions of East Antarctica, where ice is grounded deep below sea level.
That's because after the ice sheet formed, its huge and crushing weight deformed the land surface beneath it over time, allowing these particular areas to become submerged. And these regions, where ice is not on land but rather lies on a foundation deep beneath the sea, are expected to be vulnerable well before 600 parts per million of carbon dioxide levels are reached, Galeotti said.
"The part of the ice sheet now sitting below sea level is making this ice sheet even more vulnerable than the Eocene-Oligocene boundary ice sheet, because there are different mechanisms for melting," said Galeotti. Specifically, warm ocean water, or ocean currents, now have access to the ice sheet, which was not the case when it was fully on land.
Such differences between now and 34 million years ago, however, also signal a major caveat when drawing inferences linking Antarctica's ancient past to the present. After all, other processes than those associated with carbon dioxide may have further enhanced the growth of the Antarctic ice sheet back then - processes that are either no longer operative today or not really relevant because of the speed at which we are now changing the planet.
For instance, in an accompanying commentary in Science, Carolyn Lear of Cardiff University and Dan Lunt of the University of Bristol note that there have long been arguments that the tectonic widening of the Drake Passage, between South America and Antarctica, was involved in the original development of the Antarctic ice sheet, by allowing the continent to become more isolated. They now argue that this widening, by enabling the Antarctic circumpolar current and better connecting the Atlantic and the Pacific, may have also acted to help bury carbon dioxide in the ocean - which, in turn, would have simultaneously enhanced ice sheet growth.
We may be driving planetary carbon dioxide levels right now, but plate-tectonics is another matter - so there are clearly aspects of this story from the past that are not as relevant to the present.
And there's another caveat - studies of how the ice sheet changed over millions of years may or may not be relevant to the timescales that today's humans care about. What we really want to know is whether Antarctica is going to give up a major amount of ice in the next 100 years or more. You can't directly answer that question based on the new study.
Still, the research suggests that the Antarctic ice sheet can do things that our computer simulations, alone, may not capture, said Thomas Wagner, program scientist for the cryosphere at NASA, who is familiar with the new study.
"We run all these ice sheet models for things like the IPCC report and sea level rise projections. And they're great, they represent an amazing integration of mathematics, cutting edge computing, and integration of disparate fields like ice physics and oceanography," Wagner said.
"But we also know from the geologic record that the big ice sheets undergo major, rapid changes from time to time that are difficult to capture in our models but can cause sea levels to rise very rapidly - as in more than five feet in 100 years. Results like this one - connecting Antarctica to atmospheric CO2 levels not far off from where we are now - are important because they tell us what the ice wants to do. It helps guide future modeling, gives us a new way to frame our current studies, and provides bounds for hazard assessment for sea level," he said.
As is often the case with major Antarctic research, the new study is also an example of international collaboration. The scientists involved are from Italy, the United States, New Zealand, the Netherlands and the United Kingdom. The ideas were hatched when the researchers came together for a summer program in paleoclimatology at Urbino in Italy, said Galeotti.
The bottom line, said DeConto, is that the new research is "just adding to the mountain of evidence that, when greenhouse gas concentrations were high in the past, climate was warmer, there was less ice" in Antarctica. "And even in times when there were ice sheets when CO2 was higher than today, those ice sheets were variable; they grew and shrank."
The more we come to understand Antarctica's history, the more this continent, whose ice sheet is too big to even wrap our minds around, becomes relevant to our present.