Published online 27 January 2010 | Nature 463, 408-409 (2010) | doi:10.1038/463408a


Icy hunt for old air

Antarctic drilling project aims for a definitive record of climate.


Kendrick Taylor looks over the WAIS drill.Kendrick Taylor looks over the WAIS drill.C. FIRESTONE

"We're checking out history books made of ice," says Kendrick Taylor. A palaeoclimatologist at the Desert Research Institute in Reno, Nevada, Taylor is the chief scientist of the West Antarctic Ice Sheet (WAIS) Divide drilling project, which is now three-quarters of the way towards pulling up the most temporally precise record of carbon dioxide for the past 100,000 years. The highly anticipated ice core promises to improve climatologists' understanding of the dynamic global climate system, and has already begun to illuminate how humans can affect it.

On 25 January, drillers finished the season at a depth of 2,561 metres, about a kilometre off the project's final goal. Relentless winds and poor visibility at the WAIS Divide camp, about 1,170 kilometres from the South Pole, had permitted only 35 days of drilling this year.

But bad weather is precisely why researchers chose this desolate stretch of ice. Snowfall accumulation here, about half a metre per year, is an order of magnitude greater than at the sites of other Antarctic cores covering the same time span. The heavy precipitation produces annual layers of ice 22 centimetres thick near the surface, allowing palaeoclimatologists to collect season-by-season data further back in time than any other Antarctic core.

“This is the best spot on the planet to get the record we're looking for.”

The drill site is perched atop an ice divide that sends ice flowing in opposite directions. There is little lateral flow on the divide itself, ensuring that ice sampled here has not travelled in from elsewhere on the ice sheet and scrambled the climate record. "This is simply the best spot on the planet to get the record we're looking for," says Taylor. "This is where the library is."

That record should yield annual data for the past 40,000 years. It will extend back another 60,000 years, but annual layers cannot be reliably identified for that period because the weight of overlying ice has compressed them, making them too thin.

Thirty-seven investigators lead 27 projects at the WAIS Divide, funded by the US National Science Foundation, studying everything from trace elements in the ice to the potential for life deep in the core. The bubbles in the ice are a special prize. Although temperatures at the camp are well below freezing, four refrigeration units in the drilling facility keep the cores below −20 °C, the temperature at which the bubbles can leak out. Their contents are precious: they trap air from the time of snowfall, offering snapshots of ancient atmospheric conditions.

Last November, for instance, palaeoclimatologist Richard Alley and his colleagues at Pennsylvania State University in University Park reported the isotopic composition of methane extracted from the top layers of the WAIS core, representing the past 1,000 years1. They found an increase in heavy methane — a by-product of biomass burning and other human activity — around the sixteenth century, which Alley attributes to Native Americans razing forests to expand their territory as the Americas experienced a population boom. This peak recedes as quickly as it rises, corresponding to widespread deaths as Europeans arrived with their diseases and weapons. The imprint shows up in the ice record because methane circulates around the globe in a matter of years.

Yet of all the gases within the WAIS ice, Alley sees the most promise in CO2. "We're going to get the highest-resolution, best-dated CO2 record ever," he says. "That's what gets me really excited."

High-resolution ice cores have been extracted before: Greenland's GRIP and GISP2 cores, drilled in the early 1990s, have the same season-by-season detail as those from the WAIS Divide, and have been the gold standard for palaeoclimatology in the Northern Hemisphere. But dust blown in from exposed land nearby interacted with acids in Greenland's ice to produce extra CO2, which has stymied attempts to establish a reliable CO2 record there.

Most of what is known about past CO2 levels thus comes from the dust-free Antarctic ice. "WAIS will combine the high time resolution we see in Greenland with a record only retrievable in Antarctica," Taylor says.

Click to enlarge.

A question of timing

The gas record may help researchers to better understand the precise timing of past increases in CO2 and temperature. Palaeoclimatologists already know that these changes have taken place roughly in step in the past, but which rises first, the thermometer or the greenhouse gas? "All the analysis everyone has done suggests that CO2 lags temperature on timescales of several hundred years," says Ed Brook, a palaeoclimatologist at Oregon State University in Corvallis. The rising CO2 presumably acts as an amplifier to drive the temperature up further, but the margin of error in deep Antarctic cores is too large to nail down the timing of that relationship precisely. "The uncertainty is of the same order as the actual lag," says Brook. "WAIS Divide should help us solve that problem."

The CO2 record could also help to solve a recently discovered climate puzzle. Other cores have shown that past temperature changes in the Arctic are inversely coupled to changes in the Antarctic, such that hot and cold periods 'seesaw' between the poles2, 3.

One driver is the oceanic 'Atlantic conveyor belt', in which cool, salty water sinks in the North Atlantic and flows southward. Abrupt changes in heat or salinity can disrupt the flow; warm periods in the Arctic, for example, are thought to have led icebergs to dump fresh water into the North Atlantic, decreasing the salinity and density of northern seawater and preventing it from sinking. The cold water stays in the Arctic, cooling it down, while Antarctica does not receive the cold water it normally would, shifting southern temperatures up.

That, at least, is the theory. But temperature can be exchanged between the poles through other channels, including the atmosphere, which moves heat faster than ocean currents do, says Bo Vinther, a postdoc at the University of Copenhagen, who has worked on coring projects in Greenland and Antarctica. If palaeoclimatologists see long lags between temperature changes at the poles when they stack the WAIS cores up against those from Greenland, that evidence would suggest a lead role for the ocean in the bipolar seesaw. But a shorter lag would suggest a more prominent role for the atmosphere.

The CO2 record will help to answer this question, too: CO2 from deep in the ocean, where large masses of organic matter have decomposed, has a different isotopic signature from terrestrial or atmospheric CO2. Heavier CO2 in the WAIS record would suggest that ocean circulation bringing up deep water was largely responsible for the bipolar seesaw.


With about a kilometre of ice left to pull up, Taylor is hopeful that the core will be complete by the end of the next drilling season. But the final metres can be the toughest to dig up because the drill's descent grows longer with each segment of extracted core.

"We can get it all done next season if everything goes perfectly," he says. "But this is Antarctica. Every day is a surprise." 

For more images from the camp, see For a blog from the author on the rest of his trip, see andenvironment/antarcticfieldresearch_2010/">here.

  • References

    1. Mischler, J. A. et al. Glob. Biogeochem. Cycles 23, GB4024 (2009). | Article
    2. Blunier, T. & Brook, E. J. Science 291, 109-112 (2001). | Article | PubMed | ISI | ChemPort |
    3. EPICA Community Members Nature 444, 195-198 (2006). | Article | PubMed | ChemPort |
Commenting is now closed.