As effects of warming grow, UN report is quickly dated

What did take climate scientists by surprise is the accelerated melting of Arctic sea ice. Modellers have been predicting for decades that the climate change should proceed fastest above the Arctic Circle. The reason has to do with a powerful positive feedback loop known as Arctic Amplification. As the temperature rises, the ice pack covering the Arctic Ocean starts to melt, replacing white, reflective ice with darker ocean water. The water absorbs solar energy, then radiates its heat back into the atmosphere, which melts more ice, revealing more water, and so on.

That’s only part of the feedback story, says Mark Serreze, Senior Research Scientist at the National Snow and Ice Data Center (NSIDC) in Boulder, Colo. In the autumn, Arctic sea water is warmer than the air above it, but when the ice cover begins to return, it forms an insulating blanket that keeps the heat from transferring to the atmosphere. “When you lose sea ice cover,” says Serreze, “you lose that insulator.” The more open water you have in summer, the longer ice takes to re-form, so this natural insulator is absent for longer than it would normally be. A recent study by Serreze and Julienne Stroeve, a colleague at the NSIDC, showed that the rapid disappearance of Arctic summer sea ice is being amplified throughout the region, with autumn air temperatures in the last four years 3C higher than the 1978 to 2007 average.

Since the late 1970s, satellite observations have shown a steadily growing retreat of Arctic sea ice in summer. Earlier models projected that between 2050 and 2070 the north polar sea would be essentially ice-free for at least part of the year. But in 2005, a steady downward trend in summer ice started to plunge more sharply. It got worse in 2006 and 2007, and moderated only slightly this past summer. The area of the Arctic Ocean now covered by sea ice in summer is only about half as large as in 1950, according to satellite photos and data from earlier studies. The year-round sea ice is also appreciably thinner, often only one metre as opposed to three metres a half-century ago.

If that trend continues, says Serreze, “the move to ice-free will come a lot earlier, say, around 2030. Some people are even saying it could happen as early as a decade from now.”

There are some issues with the models, acknowledges Marika Holland, a climate modeler at the National Center for Atmospheric Research who specialises in sea ice. “They’re incomplete - obviously, because they’re models," she says. "You can’t include all the gory details.” Still, she says, “the fact that all of the models show a slower rate of melting than we actually see is an eye-opener. Clearly we’re missing something.” Most models, for example, don’t include the effects of soot deposition from faraway smokestacks, and tailpipes, which can darken ice and make it melt more quickly. “How important that might be isn’t entirely clear,” she says.

“Another thing about amplification,” says Serreze, “is that at some point it’s no longer confined to the ocean. You’ve got atmospheric circulation that sends heat out over land.” In the Arctic, much of that land is deeply frozen permafrost. When you warm it up, organic matter that’s been trapped in the soil for many thousands of years decomposes to release carbon dioxide and also methane - a far more potent greenhouse gas, molecule for molecule, than CO2.

Is it happening already? “The models have generally suggested we’d start to see a signal somewhere between 2012 and 2045,” says Dave Lawrence, a modeller at NCAR who specialises in the terrestrial aspects of climate change. “Some observational studies suggest that there have been significant increases in methane, maybe by as much as a factor of two. But you can’t see it from satellites, and we have a very limited number of ground stations. It could be less,” he says, “or it could be a lot more.”

Recent scientific cruises in the Arctic Ocean, however, have discovered high levels of methane bubbling up from the sub-sea permafrost, which apparently is starting to thaw as the Arctic Ocean warms.

These feedbacks are less important in the Antarctic, where most of the ice lies over land, not water, and where the land is essentially all rock, not permafrost. As a result, nobody had expected Antarctica to be warming up much yet. In fact, temperature readings at the South Pole and at Russia’s Vostok Station, in East Antarctica, have shown a slight cooling over the past few decades. That, climatologists believe, is due in part to the ozone hole that opens over the continent every Antarctic spring. At the same time, the Antarctic Peninsula, a finger of land jutting up from West Antarctica toward the southern tip of South America, has been warming rapidly, with mid-winter temperatures in the northwestern peninsula soaring more than 5C since 1950. Large ice shelves, such as the Larsen B in the Weddell Sea, have collapsed in recent years.

“People naïvely assumed that West Antarctica must be more like East Antarctica than like the Peninsula,” says Eric Steig, of the University of Washington. It’s not. In a letter to Nature last month, Steig and several colleagues used both satellite and weather-station data to show that West Antarctica has been steadily warming for the past 50 years, at an average rate of 0.1C per decade. And before the ozone hole opened up in the 1980s, East Antarctica was warming up too, and will presumably resume its warming trend as the hole repairs itself now that ozone-destroying chlorofluorocarbons have been banned.

Finally, one major source of uncertainty in projecting future climate change is what you assume about the growth of anthropogenic CO2 and other greenhouse-gas emissions. Those assumptions are based on projections of economic growth, changes in the efficiency of energy production, the rise of renewable fuels to replace gas, coal and oil, and all sorts of other factors.

Just as with the model projections of the planet’s response to greenhouse gases, the projections of emissions don’t always go according to expectations. According to a report from the National Oceanographic and Atmospheric Administration, atmospheric CO2 increased at a rate of 2.2ppm (parts per million) in 2007, compared with an average of 2ppm per year from 2000-2007, and 1.5ppm in the 1990s. Because feedback mechanisms have only lately been kicking in, most of that has to do with increased burning of fossil fuels. The trend isn’t surprising, given the growth of industry in India and China; what’s worrisome is that its magnitude is greater than that of the most fuel-intensive of the models IPCC uses in forecasting future warming.

None of this information was available when the last IPCC report was being put to bed. Neither, on the other hand, was the fact of the current recession. The plunge in economic activity has already led to a pullback in fossil-fuel use. The IPCC is just now beginning to “scope” its next major report, in the quirky language of its website, and depending on the state of the world economy when the Fifth Assessment Report goes to the printer in 2014, the forecasts will be adjusted accordingly - inevitably to be challenged as the physical world follows with more surprises.

But whatever happens, the carbon dioxide we’ve already added to the atmosphere - about 100ppm over the 280ppm that were there before the Industrial Revolution - are going to stay there. And according to a paper published in the Proceedings of the National Academy of Sciences last month, the effects on temperature, sea level rise and altered weather patterns are already locked in, even if we don’t yet know exactly what they’ll turn out to be, and are likely to persist for 1,000 years. Anything we add from today forwards will come on top of that. Just like the Fourth Assessment Report, today’s concentration of human-generated greenhouse gases, already high enough to be changing the climate for the worse, are quickly going to be out of date.