The totality of the work was published in two separate papers Monday in the Proceedings of the National Academy of Sciences. In one of them, Richard Levy of New Zealand's GNS Science, Harwood, and a large team of international colleagues presented their results from analysis of the ANDRILL-2A drill core, a more than half mile long core of ocean mud and sediment extracted near the U.S. Antarctic base McMurdo Station, as part of a multinational scientific collaboration to learn more about Antarctica's deep past.
Meanwhile, Edward Gasson of the University of Massachusetts, Amherst and four colleagues simultaneously published the results of a sophisticated new computer simulation of Antarctica that, they said, could reproduce the ice behavior implied by the massive core.
"Finally, our understanding of physical processes and our knowledge of the shape of the Antarctic continent beneath the ice -- determined through decades of data collection -- reproduce conditions that match important records like ANDRILL," said Jamin Greenbaum, an Antarctic researcher and research associate at the University of Texas at Austin who led research last year finding a key vulnerability in the gigantic Totten glacier of East Antarctica, but was not involved in the current study. "Results like these will give us added confidence in predictive models meant to forecast future sea level rise."
The ANDRILL-2A drill core itself can only be called a treasure trove of information about Antarctica's deep past. It was extracted from a floating ice platform over the ocean about 20 miles offshore, not far from the TransAntarctic Mountains, a vast range dividing East and West Antarctica.
Scientists can gather large volumes of information about past seas and climates from such an object, by studying the layers of sediment that settled over eons. These contain not only records of past animal and plant life -- ranging from the shells of tiny organisms to residues of pollen and spores -- but also isotopes of the element oxygen, which have long known to provide a telling record of the state of the planet's oceans and ice sheets.
"When the ice sheets grow, they grow with a greater amount of Oxygen 16 in them, it's the more volatile, reactive one," says Harwood. "So when you store ice on land, the oceans get enriched in Oxygen 18."
The reason is that Oxygen 16 is lighter and evaporates more easily from the ocean. Ice sheets are fed, over long time periods, by the falling of snow on top of them - snow that originates as evaporation off the seas and falls as precipitation. The more that happens, the more Oxygen 16 is stored in ice sheets, and the more Oxygen 18 is remaining in the ocean.
When it comes to learning what Antarctica -- the planet's not-quite-sleeping ice giant -- is capable of, of particular interest to scientists is a period called the Mid-Miocene climatic optimum, some 14 to 15 million years ago. Targeting this period as well as millions of years before it in the drill core, the researchers observed evidence suggesting that Antarctica's ice both advanced much farther out into the sea than where it now rests during cool phases, but also retreated much farther inland than we see at present. And seas fell, or rose, accordingly.
As it changed, the ice sheet left currently ice-crushed coastlines exposed, providing a much more hospitable environment for life plant and animal life -- and this seems to have occurred repeatedly during times when carbon dioxide concentrations in the atmosphere were at their highest. (Carbon dioxide and temperature changes at that time were not caused by humans, of course -- rather, they resulted from more slow-moving changes in the orbit of the Earth.)
"The fact that we're at 400 already, and preindustrial we were at 280, we've already gone 120," says Harwood, referring to parts per million of carbon dioxide in the atmosphere. "So we only need to go that far again until we reach this potential for the variability that we're seeing in the Miocene, in terms of the way the ice sheet can respond." The drill core, he said, suggests not only that Antarctica probably gave up the ice contained in West Antarctica during those periods (capable of raising present day seas by about 10 feet), but also much ice from the larger East Antarctica.
Granted, there's still considerable uncertainty about just how much carbon was in the air at that time. And at least one Antarctic scientist expressed skepticism about the kinds of inferences that can be made about our planetary future from a study like this.
"I think the further back you go, the larger the envelope of uncertainty in terms of extrapolating forward," said Jonathan Bamber, an Antarctic researcher based at the University of Bristol in the UK who reviewed the paper for the Post. Bamber didn't challenge the results itself, but merely suggested that the planet was so different 15 million years ago that other factors than carbon dioxide concentrations could have played a major role in what happened to Antarctica.
For instance, Bamber noted, the Isthmus of Panama, connecting North and South America, didn't exist then (though some recent research potentially suggests otherwise). That's a factor that Bamber called "quite important, because it allowed heat transport between the Pacific and Atlantic," which in turn could have changed key features of the workings of the Earth's climate.
One thing that remains very unclear from the study is how fast Antarctica can change. The time scales involved in the new research are simply too large to allow for reducing the current findings down to human terms.
"One of the things they can't really resolve particularly well is these changes, how rapid they are, if it's tens of thousands of years, or hundreds of thousands of years," said Bamber.
Still, the climate modeling study that was published at the same time as the study of the deep ocean drill core found that by adding several new processes and features to the simulation which speed up the rate of Antarctic ice collapse, it was possible to reproduce the presumed ice loss from Antarctica during the Miocene with only 500 parts per million or so of carbon dioxide in the atmosphere.
Those factors include so-called "hydrofracture" -- in which Antarctic ice shelves, which stabilize inland ice, shatter and fall apart as water pools on their surfaces, much as happened in recent memory with the Larsen A and Larsen B ice shelves -- and "cliff collapse," in which the sheer walls of ice that linger behind after hydrofracturing also crumble, due to the relative weakness of ice as a material.
Researchers like the University of Massachusetts, Amherst's Edward Gasson and Rob DeConto have been exploring including these processes in their modeling of how the ice sheets work in order to better understand their apparent changes in past eras, including several other prior warm periods besides the Miocene.
"We're adding and including new physical processes in these models, and the result of that is we've begun to be able to simulate with these models the kinds of changes in the ice sheets that the geologists see," says DeConto. He continued: "We were able to generate sea level going up and down again, on the order of tens of meters, like 30 meters, without having to go to extremely high levels of CO2."
DeConto also cautioned that the Miocene is not a perfect analogue for our planet's future - the key role of changes in the Earth's orbit during this era pretty much ensures that. After all, it's now us, not these orbital cycles, that are in the driver's seat.
Nonetheless, DeConto underscored that the new research does highlight, above all, that the Antarctic ice sheet can really respond significantly to changes in the composition of the atmosphere. "This isn't a direct analogy for the future, but it's just one more piece of evidence," he said.
