Evidence rethink puts CO2 and ancient warming back in sync

A thin layer of ice from an area of the Antarctic where ancient ice records are collected, in polarized light that reveals ice crystals. Rethinking how ice crystal formation affects ancient data collection is helping to solve an outstanding climate puzzle. © Frédéric Parrenin

A different way to dig up links between past levels of CO2 in the air and temperatures could solve a troubling question over the historical climate. Previously, data collected from long cylinders drilled from Antarctica’s ice sheet seemed to show temperatures rising hundreds of years before CO2 levels did. If ancient warming came before a CO2 rise, then the greenhouse gas seemingly couldn’t have caused the warming. Climate skeptics have used this to argue  that the CO2 we produce today isn’t causing global warming.

Now, Frédéric Parrenin at the French National Centre for Scientific Research in Grenoble and his teammates have used a different method on these cylindrical ice cores. They say that their approach shows CO2 and temperature rises happened together during the last ‘deglaciation’, when ice sheets retreated during an abrupt warming period 20,000-10,000 years ago. “This makes it possible that CO2 was actually a cause of warming corresponding to the last deglaciation,” Frédéric told me.

Scientists have been using Antarctic ice cores, and bubbles of air from the time the ice formed trapped inside, to study climate history for over 30 years. The time capsule-like bubbles show what chemicals were in the air. Meanwhile, the amounts of different forms, known as isotopes, of elements like hydrogen, carbon and oxygen in the ice reveals the temperature it formed at. And finally, scientists figure out how old the ice and bubbles are from how deep they are in the core – and that’s where Frédéric found problems.

In-depth investigation

Frédéric Parrenin cutting an ice core into one-meter-long pieces for analysis. Image © Frédéric Parrenin

The ice that traps air first forms 50-120 metres below the Antarctic surface. Before, scientists had included this depth difference when they worked out the bubble ages by simulating the change from snow to ice called ‘firn densification’. But last year Frédéric realised that the times this gave were out of sync, because the simulations misjudged how far down the bubbles are first trapped. Using this method on different cores should have produced the same timeline, or chronology, but it didn’t. “I realised that these densification models were wrong, so I decided to look for another method to estimate the chronology between CO2 and temperature,” Frédéric said.

In a paper just published in leading research journal Science, Frédéric’s team turned to a rare but stable isotope of nitrogen known as nitrogen-15. As nitrogen-15 is heavier than normal nitrogen it settles, meaning that deeper in the snow there would be more of this isotope. Scientists can therefore use how much nitrogen-15 is in the bubbles to tell how far down in the ice column they got trapped, or ‘locked-in’. The researchers could swap that information for the problematic firn densification figures in their bubble dating efforts. “The method has existed for two decades, but as it was giving different results from densification models, people thought it was wrong,” Frédéric said.

Frédéric wanted to use this method on ice from the European Project for Ice Coring in Antarctica’s ‘Dome C’ (EDC), the most detailed record of the last deglaciation. But because scientists hadn’t trusted the methods, he and his fellow European researchers wanted to show they were reliable. To do this they referred to three abrupt changes of methane level in the air in the past 15,000 years.

Cold comfort

Frédéric Parrenin’s team at the end of the 2005-2006 field season after collecting an ice core at the Talos Dome, Antarctica, which they compared to the EDC core in this study. © Frédéric Parrenin

For example, using those events, they could compare the depth difference between EDC and other ice cores calculated by firn densification models, where the errors were less serious. Their comparisons all disagreed with what firn densification would have predicted for EDC, but agreed with nitrogen-15-based predictions. “Though it has not been checked at the onset of the last deglaciation, we are confident in the nitrogen-15 method,” Frédéric said.

And while previously EDC had shown that CO2 increases lagged temperature rises by about 800 years at the latest deglaciation, Frédéric’s team found no lag using the nitrogen-15 method. “We’re saying that CO2 and Antarctic temperature vary at the same time, within 150 years approximately, so it’s completely possible that CO2 caused the Antarctic temperature warming,” Frédéric said. “We haven’t proved it, but at least the hypothesis can now be more carefully evaluated.”

In a perspective in the same Science issue, Oregon State University’s Ed Brook underlined that an Australian and Danish team revealed similar results using a different method last year. “They concluded that CO2 lagged temperature by less than 400 years on average over the entire deglaciation, and could not exclude the possibility of a slight lead,” Ed wrote.  “In many ways, these results are not surprising, given the coupled nature of the carbon cycle and climate and the fact that oceanic processes around Antarctica probably play a key role in glacial-interglacial CO2 dynamics. They are important, however because they improve our understanding of when CO2 changed with respect to temperature in the ice core record.”

Journal references:
F. Parrenin, V. Masson-Delmotte, P. Köhler, D. Raynaud, D. Paillard, J. Schwander, C. Barbante, A. Landais, A. Wegner, J. Jouze (2013). Synchronous Change of Atmospheric CO2 and Antarctic Temperature During the Last Deglacial Warming Science, 339, 1060-1063

E. Brook (2013). Leads and Lags at the End of the Last Ice Age Science, 339, 1042-1043