Climate scientists have long argued that ancient air trapped in Antarctic ice is the smoking gun that links carbon dioxide to global warming. Over the past 800,000 years or so the planet has gone through a series of ice ages interspersed with relatively warm periods (during which glaciers retreat back toward the poles) – and inevitably, these warm interludes happen when there’s more CO2 in the atmosphere.
The only tricky part of this argument is that the smoke seems to come before the gunshot. It’s most apparent in the most recent warming period, which began about 19,000 years ago: the temperature seems to begin rising before CO2concentrations increase. Climate sceptics have argued that since effects don’t come before causes, the whole theory falls apart.
In fact, it’s not much of an argument, since even a little bit of warming would release extra carbon dioxide into the air, leading to a feedback loop, causing even more warming. But whatever feeble merit the sceptic argument might have had, a new study just published in Nature pretty much demolishes it. It’s the most comprehensive analysis ever done of carbon dioxide and temperature at the end of the last ice age, and it shows quite clearly that in most of the world, the thermometer began to shoot up only after the atmosphere was spiked with carbon dioxide. “I think,” said Jeremy Shakun, a Harvard postdoctoral fellow and the lead author of the study, at a press conference, “this ends the sceptic argument.”
Shakun’s confidence is based on the comprehensiveness of the research. Most of the evidence for an ancient CO2-warming link comes from cores drilled out of Antarctica’s two-mile-thick blanket of ice. Air bubbles from different levels show how much of the heat-trapping gas the atmosphere held at different times, and the chemistry of the ice trapping the bubbles shows what the temperature was.
The problem, Shakun said, is that “these cores tell you only about temperatures in the Antarctic.” Just as you’d never infer global temperatures today from just a couple of sites, it’s not really reliable to look only to ice at the South Pole for global temperatures back then. So Shakun and his co-authors gathered no fewer than 80 different records of ancient temperatures, including lake sediments (different types of pollen at different depths point to what growing conditions were like) or sea-bottom cores (the shells of marine plankton, whose chemistry depends sensitively on ocean temperatures). It was, writes the British Antarctic Survey’s Eric Wolff in an accompanying Nature commentary, “. . . a major achievement: the difficulties of synchronizing the records and of ensuring that they are sufficiently representative of the whole planet, are considerable.”
What they found was that in Antarctica, there was indeed a bit of warming that preceded the rise in atmospheric carbon dioxide – but just a little, and only by a couple of hundred years. In the rest of the world, Shakun said, “global temperature clearly lags the CO2 buildup.” Cause, in short, really did come before effect.
The sequence of events as the authors see it is this: Around 20,000 years ago, changes in the earth’s tilt and orbit around the sun brought a little more sunlight than average to the Northern Hemisphere, where massive glaciers covered much of North America and Europe. The glaciers began to melt, dumping fresh water into the North Atlantic. Since freshwater is less dense than salt water, this slowed something called the Atlantic Meridional Overturning Circulation (AMOC) – a current that includes the Gulf Stream, and which funnels heat from the southern hemisphere to the Northern (without the AMOC, Paris, which is at about the same latitude as Fargo, North Dakota, would be drastically chillier in winter).
Since the Southern Hemisphere was no longer shipping its heat northward, Antarctica began to warm, releasing a burst of carbon dioxide into the atmosphere – somehow. “Where it’s coming from,” Shakun said, “is a big question in paleoclimate.” One mechanism is simply that warm water can’t hold as much of the dissolved gas as colder water, so any heating of the ocean would have released some, just as a warm Coke loses its fizz faster than a cold one.
Beyond that, sea ice, which would have previously formed a lid on this CO2, would have melted back near Antarctica, letting more escape. “People also think,” Shakun said, “that there would have been lots of extra carbon stored in the depths of the southern ocean. Warming could have shifted the prevailing winds, pulling up this deeper carbon and helping it degas.”
It’s a persuasive story, albeit a bit complicated, but it does leave a couple of unanswered question. One is whether water dumped into the North Atlantic by melting glaciers would have affected the AMOC by just the right amount. There’s good evidence of the melting, say the authors, both in evidence left by the glaciers themselves and in a rise in sea level that happened at the same time, but the changes to the AMOC are assumed, not proven. (We wrote about another glacier-related rise in sea level last week).
Another is that while the level of sunshine did increase in the north, it started from a relatively low point. Still another, Wolff says, is that a period of warming in the Northern Hemisphere some 60,000 years ago did not lead to an overall global temperature rise – so what’s the difference between that episode and the more recent one?
All good questions, but Shakun and his collaborators are convinced their analysis will stand up, even if all the ‘i’s’ haven’t been dotted or the ‘t’s’ crossed. “As a diligent scientist,” he said, “I never say never, but I think that while our analysis will get better as we go from 80 records to 800, it’s pretty unlikely that things will change significantly.”