On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
Climate models of the past, present and future seem to be in no short supply these days. But a new and dynamic picture of climate change appears in this week's Science, one that could affect the way future conditions are predicted.
Recent history has been kind to humans, providing a relatively stable climate for about the past 10,000 years. Many previous models have re-created short glimpses of this past.
But, says Axel Timmermann, a professor of oceanography at the University of Hawaii at Manoa, "None of these snapshots were able to capture abrupt climate change and transition," thereby making them less useful for predicting coming sudden shifts. Even if the near future doesn't unfold like the 2004 climate-gone-haywire film The Day After Tomorrow, scientists need to be able to produce accurate models of what abrupt change (more likely spanning hundreds or thousands or years, rather than days) would look like and why it might occur, explains Zhengyu Liu, lead author of the study and director of the University of Wisconsin–Madison's Center for Climate Research.
The new computer simulation takes a stab at doing just that. Based on the National Center for Atmospheric Research's Community Climate System Model, it gives scientists a novel—and, so far, reliable when checked against data from ice core samples—understanding of some causes and effects of rapid climate change.
Liu and his colleagues started their modeling at about 21,000 years ago—the zenith of the last ice age. But they've been especially interested in the most recent period of abrupt global warming, the Bølling-Allerød, which occurred about 14,500 years ago when average temperatures in Greenland rose about 15 degrees Celsius in about 3,000 years. The causes of the warming remain debated, but Liu and his team homed in on the melting glacial water that poured into oceans as the ice receded, paradoxically slowing the ocean current in the North Atlantic that keeps Europe from freezing over. Should ice in Arctic regions, such as Greenland, melt again, the globe may face a similar situation, sending Europe into a colder age despite warming taking place elsewhere.
Behind the team's process is a diverse group of experts (from geology to oceanography to computer science), a wealth of hard data, and some powerful supercomputers. Active elements of the model include the atmosphere, ocean, land surface—including vegetation—and ice. It also included prescribed information for solar radiation and carbon dioxide levels.
With the immense amounts of data and exhaustive supercomputer crunching, Liu says, "we have to be very careful." The model requires so many calculations that it could only be run once—even with more than a year of computer time at the Oak Ridge National Laboratory, in Tennessee.
The team's simulation has now reached to about 10,000 years ago and is still plugging away at the modeling, working its way toward present day, Liu reports, shoring up the results with ice core and other dug-up data along the way.
"The general idea is to test for the present and past and hope that [the model] works in the future," says Timmermann, who wrote an accompanying perspective in the Science issue. "This is a wonderful step forward."
