Science, October 5, 2007


[Rachel's introduction: "This [arctic] ice is more vulnerable than we thought." And that vulnerability seems to be growing from year to year, inspiring concern that Arctic ice could be in an abrupt, irreversible decline.]

By Richard A. Kerr

A few years ago, researchers modeling the fate of Arctic sea ice under global warming saw a good chance that the ice could disappear, in summertime at least, by the end of the 21st century. Then talk swung to summer ice not making it past mid-century. Now, after watching Arctic sea ice shrink back last month to a startling record-low area, scientists are worried that 2050 may be overoptimistic.

"This year has been such a quantum leap downward, it has surprised many scientists," says polar researcher John Walsh of the University of Alaska, Fairbanks. "This ice is more vulnerable than we thought." And that vulnerability seems to be growing from year to year, inspiring concern that Arctic ice could be in an abrupt, irreversible decline. "Maybe we are reaching the tipping point," says Walsh.

There's no doubt that 2007 was a special summer melt season. The ice area remaining in September--the year's low point--had been shrinking since satellite monitoring began in 1979. Some years it recovered a bit, others it declined further, but overall it shrank 8.6% per decade. In 2005, it hit a record low of 5.6 million square kilometers, down 20% from 1979. But last month, "we completely blew 2005 out of the water," says sea ice specialist Mark Serreze of the University of Colorado, Boulder. Ice area plummeted to 4.13 million square kilometers, down 43% from 1979. That's a loss equivalent to more than two Alaskas. The new low is more than one Alaska below the trend line. Nothing else like that appears in the satellite record or, for that matter, in monitoring from ships and planes during the rest of the 20th century, says Walsh. An immediate cause of the record-breaking year is clear enough. As Serreze explains, an unusually strong high- pressure center sat over the central Arctic Ocean while a strong low hovered over Siberia. This weather pattern allowed more solar heat through the clear skies beneath the high-pressure center and pumped warm air up from the south between the high and the low.

The vicissitudes of weather may have enhanced ice loss this year, but there's more going on than that, scientists are realizing. For one thing, their models underestimate how fast summer ice has been disappearing in the warming Arctic. "It's very alarming the way things are changing so fast," says polar oceanographer D. Andrew Rothrock of the University of Washington (UW), Seattle. "We've thought we have the important physics in the models, but... it seems our models aren't very good in the Arctic."

Researchers say the models probably lack some realistic feedbacks, natural processes that can amplify a climatic nudge--whether natural or humanmade--into a shove. And that shove could send the ice past a tipping point. "You get a kick in the right direction," says Serreze, "and it sends the ice over the edge" and into a meltdown from which it cannot recover.

Last December, researchers reported finding that at least one climate model includes feedbacks that can accelerate sea ice into a tipping point. Modeler Marika Holland of the National Center for Atmospheric Research (NCAR) in Boulder, Colorado, and colleagues wrote in Geophysical Research Letters (GRL) that when NCAR's Community Climate System Model, version 3--which has one of the most sophisticated ice components available--is run under a strengthening greenhouse, sea ice loss can suddenly accelerate, in one case cutting ice area by twothirds in a decade and wiping out September ice by 2040.

Such accelerations were driven by two feedbacks in the model. In one, thinner ice one year made ice melt more easily the next year. In another, when white, highly reflective ice melted, the darker, more absorptive open water that replaced it absorbed more solar energy. The added heat could help melt more ice and keep new ice thinner that year--and even the next, if the heat lingered through the winter. Holland and her colleagues "showed that in models, these abrupt changes can occur," says Walsh. Now, "this is the first time we may have seen it" in the real world. In an in-press GRL paper, polar researcher Donald Perovich of the U.S. Army Cold Regions Research and Engineering Laboratory in Hanover, New Hampshire, and colleagues report estimates of increasing solar heating of the Arctic Ocean. They found that a large area of Arctic waters north of the Bering Strait had been absorbing increasing amounts of solar heat since 1979 as summer ice retreated, suggesting that the ice-reflectivity feedback has been operating there.

And in a paper appearing in GRL this week, Son Nghiem of the Jet Propulsion Laboratory in Pasadena, California, and colleagues report a continuing decline in the thicker, older ice that tends to persist from year to year. Much of the decline in perennial ice, they found, was due to winds blowing it out of the Arctic Ocean. But thinning from added heat had made it easier for the wind to blow the ice out. That would add a dynamical feedback to the thermal feedback of ice reflectivity.

Researchers suspect that these and other feedbacks are eroding sea ice's ability to resist the warming of recent decades. "Might we lose summer sea ice by 2030?" asks Serreze. "That is not unreasonable." Next September could tell whether natural variability just made for one bad year in the Arctic or whether it is pushing the ice over the edge. Meteorologist Ignatius Rigor of UW is wor ried. Given the beating the ice has taken of late, he says, "the chances of another extreme next year are pretty high."

Copyright 2007 American Association for the Advancement of Science