91av

In the waters around the warming continent, the icy grip of winter seems stronger than ever. We explore the mystery of the Southern Ocean sea ice

The world is set to defrost. All over the planet glaciers are retreating, while tundra thaws. The ice caps of Greenland and Antarctica are looking fragile, and the Arctic’s once-vast raft of sea ice is shrinking at an alarming pace. And down south, in the seas around Antarctica, the sea ice… well… er… seems to be growing.

In the few decades we have had satellites keeping watch, the area of the Southern Ocean covered by sea ice in winter has grown bigger, hitting record levels in recent years. The increase is small, but it is surprising – and something of a mystery. “The Arctic is doing exactly what we would expect,” says Paul Holland of the British Antarctic Survey. “The Antarctic is not.”

A couple of years ago, Holland thought he had cracked the mystery: stronger winds were to blame, his team concluded. But now Holland thinks they got it wrong. So what in all the icy hells is going on down there? The answer matters to more than penguins. If or when the sea ice starts to shrink instead of growing, it could have knock-on effects around the world, from interfering with ocean currents to giving a nudge to the teetering West Antarctic ice sheet, whose collapse would raise sea level by several metres.

Why is the ice increasing? (Image: Caul Nicklen/NGS)

Locked up

Vast amounts of water are locked away in the great ice sheets in Antarctica. Almost the entire continent is covered by ice sheets that are nearly 2 kilometres thick on average, hiding whole mountain ranges. Where this ice slips off the land into the sea, it forms floating ice shelves hundreds of metres thick. Half the coast is surrounded by ice shelves, some of them immense.

Beyond lies the sea ice. It is distinct from the ice shelves, because it forms when the surface of the sea freezes and is about a couple of metres thick on average. Unlike in the Arctic, in Antarctica almost all the sea ice melts in spring and reforms each autumn and winter (see “Polar opposites”, below).

Now the frozen continent is warming up – with unexpected consequences. As recently as 2007, the official prediction was that the ice sheets would grow over the 21st century, because higher snowfall would more than compensate for higher ice losses. In reality, satellite gravity measurements show the ice sheets have already started to shrink.

Ice shelves are not following the script either. They have been thinning faster than expected, and . The collapse of the enormous Larsen B ice shelf in 2002 shocked most glaciologists.

Most baffling of all is the behaviour of the sea ice. According to the majority of climate models, it should be shrinking as the air and waters around Antarctica warm. And in some places, such as in the Amundsen and Bellingshausen seas west of the rapidly warming Antarctic Peninsula, the sea ice is doing just that. But in others, it is growing (see “Thick and thin”). Overall, the area covered by sea ice in winter is slowly increasing.

This is good news. Although Antarctic sea ice is mostly around in the 24-hour darkness of winter, there are still several million square kilometres of it left in the spring when the sun is high. By reflecting a little more sunlight, the extra spring sea ice should slightly slow the warming of the seas around Antarctica. (In contrast, the Arctic Ocean is absorbing more heat and warming faster as the area of summer ice shrinks – a positive feedback.)

Indeed, satellite measurements show that the oceans around Antarctica are reflecting 0.9 per cent more sunlight in summer, says Norman Loeb of NASA’s Langley Research Center in Hampton, Virginia, whereas . The big question is what happens next. “If instead it were shrinking like the Arctic, you would imagine ,” says Holland.

The loss of Antarctic sea ice would not only lead to more warming, it would also affect ocean currents. As the sea ice forms in winter, is left behind. This cold, dense water sinks down to the ocean depths and flows around the globe before eventually slowly surfacing again in tropical seas.

The waters around Antarctica, though, are becoming fresher and less dense because the ice shelves are melting faster, and more snow and rain now fall on the Southern Ocean. This is hindering the sinking process. If the sea ice retreats then sinking might stop altogether, changing ocean circulation around the globe. That would affect sea temperatures and thus the climate in ways that are hard to predict.

So the sea ice is important, but right now we have no idea whether it will continue to slowly increase over the coming decades, or suddenly disappear. “There is a pressing need to understand this,” says glaciologist Sharon Stammerjohn of the Institute of Arctic and Alpine Research in Boulder, Colorado.

“There is a pressing need to understand what’s happening to the sea ice”

In 2012, Holland thought his team had the solution. “We made the simple claim that stronger winds from the south were carrying cold air off Antarctica, and dragging ice north,” he says. In autumn and winter, this would create gaps where new ice could form as well as cooling exposed water. A of satellite images seemed to support the idea.

Blowing in the wind

It seemed like a satisfying explanation for the sea ice paradox, given that there is no doubt that the winds around Antarctica have been strengthening (91av, 20 July 2013, p 34). This is partly due to global warming and also partly the result of the hole in the ozone layer created by our pollution. Natural variability may also play a part.

But when Holland looked again at what was happening, he began to doubt the wind explanation. This time, he looked at ice changes in a different way. Instead of focusing on the area of sea ice, he looked at how fast it was melting or forming. This is a more direct way to see the influence of climate changes, Holland says. “For example, warming wouldn’t directly decrease the amount of ice in a season, but rather its melting rate.”

From this viewpoint, there is no longer an autumn lockstep between wind and ice. The most glaring clash is in the Bellingshausen Sea. There, the autumn winds have become stronger. They blow from the north, which must be shoving more ice in towards the coast, keeping it from spreading, as well as carrying warmer air from temperate regions. Indeed, the overall area of autumn sea ice here has declined, which would seem to fit with that idea. But if the stronger winds were the key factor affecting ice growth, the ice should be growing more slowly. In fact, during autumn it is growing faster than it did a few decades ago. “That destroys my earlier work,” Holland says.

He now thinks that to understand these changes in sea ice, we need to focus on what’s happening in the spring. In the Bellingshausen Sea, ice is now retreating earlier in the springtime, letting the ocean absorb more solar heat. That warming should delay the regrowth of ice – accounting for the decline in ice area in autumn. But when the ice does grow back, it does so quickly, because now the ocean is open to the air, it rapidly loses its heat again.

In other words, sea ice has a tendency to bounce back from big spring losses. This was first noted in Antarctica by Stammerjohn, in at how ice is changing region by region and season by season.

These feedback processes also happen in the Arctic, says Dirk Notz at the Max Planck Institute for Meteorology in Hamburg, Germany, which may explain why summer sea ice there seems to recover somewhat every time it hits a new record low. None of this even begins to explain why Antarctic sea ice is growing overall, but if these trends originate in the spring, that’s where we should look for what’s really causing them, says Holland. “I hope it is a trail that will lead to the truth.”

Notz, however, is not convinced that Holland is right to focus on whether sea ice is melting or forming faster or more slowly than it used to. “He is looking at a change in a rate of change,” he says. “I do not think it’s a measure that is important.”

So Notz’s team thinks that the wind explanation still holds. The reason most climate models have been projecting sea ice losses, , is that they are too coarse-grained. They miss details of Antarctic topography that deflect winds northwards and spread out sea ice, allowing more ice to form.

However, there are almost certainly other forces at work. Changes in deep ocean circulation are bringing in more warm water around West Antarctica. This is thought to be the main reason for the thinning of the ice shelves there, says Stammerjohn, and may be speeding up the loss of sea ice too. Elsewhere, she suggests, less ocean heat may be welling up, allowing more sea ice to form.

And could the rush of fresh water from the thinning ice shelves be playing a part, too? Fresher water not only freezes more readily, it is also more buoyant, so a surface layer of fresher water .

However, the greatest increase in fresh water is around the Amundsen Sea, where the glaciers are retreating fast, yet sea ice there is still shrinking, too. So the fresh water appears to be having little effect, Stammerjohn says, although the case is not entirely closed.

Blind satellites

In the end, it is not so surprising that we are struggling to understand Antarctica. This region is a lot more complex than the Arctic, yet observations are much scarcer because the region is so remote and forbidding. Simply building instruments tough enough to survive the conditions is difficult, let alone deploying them. Even satellites see less here. In the Arctic it is possible to use radar altimeters to measure ice thickness, but in the Antarctic there is a lot more snow sitting on the ice, which absorbs the radar signal. Yet monitoring ice thickness is critical for understanding what’s happening. Robot subs are now being used to rove under the ice to measure its thickness, but so far they provide only a snapshot of a small area.

Besides better observations, we need better models. Trying to build climate models that match what’s happening in Antarctica may be the most productive way to resolve the debate about the causes of the sea ice increase. “It’s a lot of things working against or for each other, which makes it hard to get one’s head around what really will happen,” says Notz. But if he is right about the role of small-scale topography then we are nearly there – the key will be improving model resolutions from 100 kilometres or so down to a few kilometres to get the wind directions right.

Holland thinks we’re still far from the answer, but he too thinks better modelling is the way forward. “When we get a model that matches what happens in the spring, we can look in the model to see what it’s doing.”

In the meantime, with many of the other effects of global warming kicking in much sooner and harder than we expected, let’s keep our fingers crossed that the stubborn seas around Antarctica continue to buck the trend for a few more decades. Sometimes it’s good to be wrong.

Leader “Hottest year on record ramps up the climate pressure“

This article appeared in print under the headline “Thaw point”