Quotes of the Day

For scientists studying the health of Arctic sea ice, satellite observations are absolutely essential for providing the big picture. It was satellites that revealed in September 2007 a record minimum ice coverage in the region — the result of a massive summer melt. And it was satellites that showed in 2008 and 2009 the modest recovery of late-summer Arctic ice that suggested to some that the specter of a totally ice-free polar ocean might be somewhat less imminent than feared.

But those high-altitude observations need occasional reality checks from scientists down on the surface. It was during one such on-the-ground research expedition last fall that David Barber, an Arctic climatologist at the University of Manitoba, got an unwelcome surprise.

Barber was aboard the Canadian research icebreaker Amundsen, checking on ice in the Beaufort Sea north of Alaska and Western Canada. The ship was well inside a region the satellites said should be choked with thick, multiyear-old ice. "That's pretty much a no-go zone for an icebreaker of the Amundsen's size," says Barber. But the ship kept going, at a brisk 13 knots — its top speed in open water is 13.7 knots — and even when it finally reached thick ice, he says, "we could still penetrate it easily."

In short, as Barber and his colleagues explain in a recent paper in Geophysical Review Letters, the analysis of what the satellites were seeing was wrong. Some of what satellites identified as thick, melt-resistant multiyear ice turned out to be, in Barber's words, "full of holes, like Swiss cheese. We haven't seen this sort of thing before."

The findings add another wrinkle to a problem climate scientists have been warning about since the record melt of 2007: after each summer meltback, the Arctic Ocean refreezes completely in winter. The problem is that much of that refreezing creates a relatively thin layer of so-called first-year ice. "It's weaker than thick, multiyear ice," says University of Colorado scientist James Maslanik, "and less resistant to melting."

In 2008 and 2009, scientists feared the record-thin Arctic ice cover might melt away. But it didn't, because of unusually favorable ocean currents and weather patterns. "Early in the 2009 season it looked like we might be on the way to a record melt," says Julienne Stroeve, a research scientist at the National Snow and Ice Data Center, in Boulder, Colo., "but then winds spread the ice out, so the overall coverage ended up being greater than in 2007." Without those winds, in other words, 2009 might have set a new record for open water. But as it happened, ice cover in 2008 and 2009 rebounded significantly — but perhaps deceptively so.

What Barber's expedition further discovered was that some Arctic sea ice is not only whisper thin, but that even in places with thick ice, the ice was not as solid as satellites had indicated. That thick ice was still there, but largely as individual chunks covered with a veneer of new ice that masked their true nature. "It's significant and it's surprising," says Maslanik. "I wouldn't have expected that the ice would be as rotten and weak as what David Barber's team found."

This suggests that the analysis of satellite observations might be due for updating. "The algorithms we use to monitor ice extent were developed a long time ago," says Stroeve, "based on what 'typical' ice looked like at that time. We know there are errors with the measurements." The weakness in multiyear ice also suggests that if the unfavorable winds and currents that caused the 2007 meltback should recur, the Arctic Ocean could undergo another especially dramatic summer melt. Not just the first-year ice might go, but also some of the "rotten" multiyear ice that Barber encountered.

That in turn would trigger one of the many positive feedback mechanisms that could speed up the warming effects of greenhouse gases. Open ocean reflects less of the Sun's energy than ice does, so a large-scale summer melt would mean more absorption of heat in the ocean. The warmer ocean would heat the air above it, which would slow the refreezing of ice in winter, which would in turn become even more susceptible to melting in summer.

Exactly when a catastrophic melt might occur, however, is unpredictable. The long-term rise in global temperature as a result of greenhouse-gas emissions is overlaid with natural, year-to-year variability in all sorts of interconnected oceanic and atmospheric cycles that slow down warming down or speed it up temporarily. But because these variations tend to be cyclical, the "perfect storm" of conditions that caused the record 2007 melting — a situation Stroeve calls "unusual, but not unprecedented" — will probably return at some point. If they do, the Arctic could be primed for major, even irreversible, changes.

Lemonick is the senior science writer at Climate Central.

• Michael D. Lemonick
• For scientists studying Arctic sea ice, satellite observations are essential. But high-altitude analyses can miss crucial information gathered on the ground.