The Mystery at the Heart of This Year's Record-Setting Arctic Ice Melt

You have probably heard that the Arctic has less sea ice right now than humans have ever recorded. The new record, set yesterday, beat the previous low, which was measured in September 2007.

"By itself it's just a number, and occasionally records are going to get set," said National Snow and Ice Data Center scientist Walt Meier in an official statement. "But in the context of what's happened in the last several years and throughout the satellite record, it's an indication that the Arctic sea ice cover is fundamentally changing."

There are two odd things about this sad record of global change.

First, it's only late August, several weeks before the traditional time when the sea ice melting stopped. That could mean that the melt is stopping earlier and could begin to recover earlier. Or we may have several weeks to go of melting, in which case, this year's low could not just break but shatter 2007's record.

Second, if the melt continues for days or weeks more, the melt will end up catastrophically lower than anyone anticipated.

After 2007's low -- which scared many Arctic scientists into statements like, "The Arctic is screaming" -- the sea ice up north recovered, though not to pre-2007 levels. Counting this year, the six years with the least sea ice on record all occurred in the last six years.

In that sense, it is not a monumental surprise that 2012 did not see an overabundance of sea ice or return to the norms of earlier this century. On the other hand, the catastrophic drop off of sea ice in the last few weeks was not something that was easy to model or predict.

What happened?


Each year, a program called the Sea Ice Outlook gathers predictions about sea ice from different teams of scientists. The groups use different techniques to predict what the next summer's ice melt season might look like. Some do straight statistics, others build models, and other groups use heuristics or a combination of methods. Teams can submit in June, July, and August. In all cases, they are all trying to guess how much sea ice will be left in the Arctic come mid-September, which (as noted above) had traditionally been the low point. The measure that scientists use to describe the sea ice's extent is the millions of square kilometers of ice that our satellites can see from orbit.

In May of this year, it looked as if the Arctic was going to have a year much like the past four, if a little worse. A lot of ice would melt, more than in any of the years before or since 2007's record-setting low, but none predicted a catastrophic year. The median guess was 4.4 million square kilometers of sea ice would be left, and the band was pretty tight around that number, with only a single group predicting a sea ice extent of 4.1 million square kilometers or lower. A few numbers for comparison: The average low from 1979 to 2000 was 6.7 million square kilometers. In 2010 and 2011, 4.9 and 4.6 million square kilometers of sea ice remained in September. The record low was 4.17 million square kilometers of sea ice in 2007.

And it is this last number that the Arctic crashed through this week with a measurement of 4.1 million square kilometers of sea ice left.

Now, the modelers may not turn out to have been wrong if, as we mentioned above, the sea ice melt season ends early. "We'll see how September turns out," Robert Grumbine, an ice modeler with the National Oceanic and Atmospheric Administration who worked on three of the predictions, told me. "It isn't unheard of for the annual minimum to be about now, rather than in September."

Some of the more recent (July and August) predictions for the Sea Ice Outlook have revised their estimates downwards on the basis of the early summer. Nonetheless, it still stands: going into this summer, we were not expecting a record low for sea ice.

And now here's the most recent data.


What's befuddling about 2012, relative to 2007, is that the Arctic has not seen the kind of ice-melting weather that 2007 did. "I'm at a loss at this loss," wrote sea ice blogger (yes they exist!) Neven Acropolis. "The 2007 record that stunned everyone, gets shattered without 2007 weather conditions."

Unknown unknowns aside, one frightening possibility exists. For years, scientists have been warning that Arctic sea ice is thinning. Thinner ice melts more quickly. But when we traditionally measure sea ice, scientists aren't looking at the mass of the ice, just the surface extent.

Unfortunately, as the University of Washington's Polar Science Center explains, we can't monitor sea ice volume easily.

Sea ice volume is an important climate indicator. It depends on both ice thickness and extent and therefore more directly tied to climate forcing than extent alone. However, Arctic sea ice volume cannot currently be observed continuously. Observations from satellites, Navy submarines, moorings, and field measurements are all limited in space and time.

So, scientists have to take these limited readings and build a model that estimates the sea ice volume. They call it PIOMAS, a catchy acronym for the Pan-Arctic Ice Ocean Modeling and Assimilation System. Obviously, as with any model, there are limitations to its predictive power. But in recent years, and even more dramatically in recent weeks, PIOMAS has been showing a major anomaly relative to the longtime average in ice thickness. Even where there has been ice in the Arctic, it's much thinner than it used to be.


If this year's melt continues and the amount of sea ice in the Arctic reaches even lower, it may be confirmation that the ice really is that thin and easier to melt, perhaps introducing other dynamics that seem poised to accelerate the decline of the system. The climate change signal, in other words, is growing stronger as the Arctic moves toward ice-free summers. Or here's Jennifer Francis of Rutgers, as quoted on Dot Earth:

The fact that the ice is so dramatically thinner now than it was only 20 years ago means that it is vulnerable to any abnormal weather event or fluctuation in ocean currents. If the "perfect storm" of atmospheric and oceanic conditions that led to the 2007 record, or the patterns that reduced ice this summer, happened back in the thick-ice era, sea-ice loss would not be making headlines. Following this summer's new record ice loss, the Arctic will enter a winter with even less ice than ever before, leading to even thinner ice, which barring any monumental external events like a major volcanic eruption, will likely perpetuate the trend in sea ice decline.

We are all trying to understand a system that has entered territory never seen before, and I think scientists are naturally cautious in their interpretations and predictions of what's to come. Models were formulated with understanding of the familiar system, which may explain why some model simulations lag observations in losing ice.