After the Ice

by Alun Anderson

  Heat from the Pacific water slows the formation of sea ice in winter, Shimada thinks, so that it ends up much thinner than that in the surrounding area. When summer comes, it is the first to disappear. That thinning of the ice triggers feedback loops. Winds can much more easily blow along a mixture of ice and open water. If the ice is really thick, winds can’t do much more than whistle over its top. “If the ice moves faster, more warm water is drawn into the Arctic, which stops the winter thickening of ice,” explains Shimada. “The thin ice then moves faster, drawing in more water and so on.”

  Sounds simple: warm water, thinner ice, winds blow the ice more easily, warm water is carried along to just where the ice is vanishing. It is not the total volume of water coming in through the Bering Strait that is important. That has not changed, explains Shimada. But the movement of the ice is growing faster, taking the warm water farther into the Arctic and up to shallower levels. This is, however, a controversial idea, with some strong supporters, mostly oceanographers, and some who think it might perhaps be a small effect in a much bigger picture. “Most atmospheric scientists don’t want to understand my view,” says Shimada with a sigh, thinking of that divide between those who look upward and those who look downward for explanations.

  The Arctic ice itself, of course, does not give a hoot for these academic divisions, as almost anything that happens to the ice will quickly set in motion more changes, which set in motion yet others. Warm water from the Pacific might well slow the growth of winter ice and leave thinner ice that can be blown around more easily. That in turn will leave more patches of open water. And as we are about to find out, there is nothing more dangerous to an ice floe sitting out on the Arctic seas in the middle of summer than a patch of open water nearby. To wrap up the story of the vanishing ice, it is time to surface from below the seas and look at the lethal combination of sunshine and open water.

  Chapter Six

  THE LETHAL MIX

  Sitting out on top of the ice in the middle of the Arctic Ocean are six odd-looking buoys. Each of them is about the size and shape of an automobile tire and is connected to a tangle of cables that snake out across the ice. One cable connects to a short antenna sitting alongside the buoy, another disappears down a hole in the ice, and a couple of others climb up a pole that stands nearby and looks a little like a streetlight that has landed by accident in the middle of the Arctic.

  Despite their odd appearance it is easy to grow fond of these buoys. They have their own Web site where they each post their daily reports; here you can read how far they have drifted across the sea and what they have seen out there all alone in the middle of the ice. Occasionally, they come to a bad end. The six buoys sitting out there in the spring of 2009 were the survivors of a contingent of nine. The last to disappear was Buoy 2008E which “melted out” on November 14, 2008. That is, it disappeared into the sea in the vicinity of the North Pole and its data page went blank. The remaining six are still in touch and their data pages tell some good stories. These past few years, some have been seeing the emergence of dramatic feedback effects that are the last element we need to understand the great melt.

  These “mass balance buoys” are put out on the ice by the U.S. Army Cold Regions Research and Engineering Laboratory (CRREL), which has its headquarters in Hanover, New Hampshire. Each of the buoys carries a pair of “pingers,” one up on that pole standing close to the buoy and another hidden away beneath the ice. The pingers, or acoustic rangefinders, send out pulses of sound and record how long they take to echo back from the top and bottom surfaces of the ice. That gives a measure of the thickness of the ice and, over time, will show how quickly the top and bottom of the ice are melting away. Other instruments record the air and water temperatures.

  Jacqueline Richter-Menge, an engineer at CRREL, has been looking after these buoys out on the ice since 2000 and knows them well. In the summer of 2007, the year of the ice’s catastrophic disappearance from the Beaufort Sea side of the Arctic, she and her colleagues began to receive some startling messages from one of her buoys.1

  The ice on the bottom of the floe beneath the buoy was melting away at an astonishing speed. The buoy was sitting on top of multiyear ice, usually around 3.2 meters (10.5 feet) thick, built up during the winter when air temperatures had hit-45°C (-49°F). In previous summers, when the warmth of August came around, one centimeter (0.4 inches) or so of ice would melt from the bottom of the floe each day. In 2007, the melt was averaging four centimeters a day (1.6 inches), and reached a peak of eleven centimeters (4.3 inches) a day in late August. Overall, 2.1 meters (6.9 feet) of bottom ice melted away in the summer of 2007, more than six times the annual average of 0.34 meters (1.1 feet) for the 1990s and two and a half times the melt seen during the previous summer. In contrast, the amount of melting from the surface of the ice was much the same as in previous years. Mass balance buoys in a different part of the Arctic, over near the North Pole, reported nothing unusual. “We were very surprised to see so much bottom melt in the Beaufort Sea,” Richter-Menge says, “and we were suspicious about the amount of increased solar heating to the ocean too.”

  The summer of 2007 had been exceptionally sunny in the Beaufort Sea. Not surprisingly, Richter-Menge and her colleagues began their search for the cause of this high-speed bottom melt by calculating the impact that sunshine might have had on the ice and sea. The impact of sunshine can be counterintuitive. You might expect bright sun and clear skies to be lethal for ice, but it isn’t. White surfaces have a high albedo, that is to say they are good at reflecting light, while darker ones have a low albedo and reflect little light. “If you are talking about melting ice,” says Richter-Menge, “then having bright sunshine and clear days is great [that is, great for leaving the ice as it is] because the ice is a bright, high-albedo surface that reflects the sunlight back.” The opposite is true for water. Seawater is a dark, low-albedo surface which soaks up sunlight and warms up fast. With open water soaking up 93 percent of the sunlight that reaches it and pure white sea ice with a layer of snow absorbing only 15 percent of the sunlight, it is open water that is really critical in warming the Arctic.

  Richter-Menge and her colleagues collected data on the changing area of ice-free water and the amount of sunshine in the region. Putting those together, they could work out how much extra solar energy had been soaked up by the waters out in the Beaufort Sea. The answer was astonishing: the 2007 solar heat input in the area where their buoy was sitting was 400 to 500 percent higher than usual, mostly due to the extra areas of ocean able to soak up sunshine. The sunlight hitting the top of the ice had not had much impact. But as the warmer water washed under the ice floes, it melted them away.

  The extra heat soaked up by the oceans was more than enough to account for the rapid bottom melt seen by the mass balance buoy in the Beaufort Sea ice in 2007. But what is really important is that as the ice melted, the area of dark open water grew even larger. That meant yet more solar heat was absorbed, which resulted in more ice melting, round and round in a lethal feedback loop with ice as its victim. The temperature of the sea surface just north of the Chukchi Sea rose to 4°C (39.2°F) as a huge area of ice vanished.2 Usually the seawater here is at-1°C (30.2°F). The remaining ice in this area had no chance.

  The impact of the ice-albedo feedback loop was helped along by that year’s exceptionally bright, sunny weather and an unusual pattern of winds.3 Overall the conditions in 2007 added up to what Richter-Menge describes as “the perfect storm,” precipitating the catastrophic loss of ice in the summer of that year. The winds helped push ice away from the western Arctic, exposing ever more open water to soak up the bright sunshine. The pattern of winds was not a return to that positive phase of the Arctic Oscillation that had helped push ice out of the Arctic back in the 1990s. Another odd pattern has arrived instead. Jean-Claude Gascard, leader of Europe’s DAMOCLES Arctic research project, explains: “We have started to see a drastic change in the atmospheric circulation pattern. Based on the last th
ree or so years, we are seeing a pattern with a low pressure system over Siberia and high pressure system over North America. That leaves a corridor in the center of the Arctic going from the Bering Strait on the Pacific side to the Fram Strait in the Atlantic, driving a lot of air through it.” Strong winds blew straight across the Arctic. Their impact could be picked up by tracking the movement of buoys left out on the ice as part of the International Arctic Buoy program. The ice was pushed over toward the coast of Canada and Greenland, 4 leaving patches of open water behind for the bright sun to heat up. Some of that ice ended up in the Transpolar Drift Stream and vanished out through the Fram Strait.

  The catastrophe of 2007 was under way, but it would never have happened if the Arctic had still had lots of its thick, old ice. Thin ice can be pushed around the Arctic by the winds much faster than can thick ice. (Remember how the Tara repeated the epic journey of the Fram in less than half the time the Fram took; that was because the ice is much thinner now and flowing faster with the winds.) Thin ice can also be pushed together and piled up in great crumpled piles much more easily than can thick ice. The winds of 2007 were able to sweep the ice away, leaving enough open water for the ice-albedo feedback to run away in the sunny summer, because it was so thin. As the ice was thinner than it had been before, it melted into water that much faster, speeding the feedback loop ever more.

  Add together thinning ice, changing wind patterns, the albedo feedback effect, the warmth of the oceans creeping up from below, and the way they all interact with one another in a cascade of disasters, and we can try a sketch of the recent history of the Arctic ice.

  The story goes like this. The slow rise in Arctic air temperatures helped gradually thin the ice over many decades. But it may not have acted alone. Warmer water flowing into the Arctic from the Atlantic during the 1980s and 1990s may have helped thin the ice from below in the eastern Arctic. Warm water from the Pacific side is slowing winter ice growth on the west. In the late 1980s to early 1990s, the unusual patterns of winds (the positive phase of the Arctic Oscillation) helped push old ice out of the Arctic for seven years on end, leaving more of the Arctic with ice that was young, thin, and weak.

  Winds, warmer air, and warmer seas added up to leave thinner ice, 5 and with less of that really strong multiyear ice, the frozen ocean became ever more vulnerable. The ice was traveling toward the disaster that arrived in the summer of 2007. That year the new pattern of winds pushed ice out of the way and bunched it up, clearing enough open water to trigger the ice-devouring albedo feedback. Months of sunny weather powered its destructive force. The “perfect storm” hit the ice of the western Arctic so hard that its area fell to a record low.6

  It is only a sketch, but one that can neatly accommodate what happened the following year, which was nothing special at all. In the summer of 2008 nothing unusual pushed the ice farther down the road to oblivion; nor were there any exceptional events that might have helped the ice recover. It was not especially cold, nor were there wind patterns that would have trapped ice within the Arctic. Still, the area of ice left at the end of this average summer was only slightly greater than the year before, the year of the perfect storm. The ice had become too thin and vulnerable to make a comeback even when the weather was benign.

  “The media didn’t like that,” says Mark Serreze, director of the U.S. National Snow and Ice Data Center and one of the world’s best-known experts on the changing Arctic ice. “They didn’t like it because it didn’t break a record. The ice recovered a little bit. But you can look at that in two ways. The other way of viewing this is that even though the atmospheric pattern of the summer was not nearly as favorable to loss of ice as the previous year, we still went all the way to the second lowest ice area on record. That is frightening.”

  The ice may have passed a tipping point.7 Once the ice is too thin and there is too much open water, the ice-albedo effect will kick in quickly every year and melt ever more of the vulnerable ice, leaving warmer water behind to slow the growth of winter ice. Past the tipping point, the rising temperatures due to global warming (what scientists call the “external forcing”) scarcely matter anymore because the ice’s own self-destructive dynamics have taken over.

  With that sketch of what has been happening to the ice, it is possible to start following what is happening in scientists’ minds. In a unique project scientists have been putting their thoughts on public display. In 2008, the U.S. SEARCH (Study of Environmental Arctic Change) program began asking sea ice research groups around the world to send in a monthly “outlook” giving their views on the sea ice and, if they dared, a prediction of the area of the sea ice at the end of the summer. All the ideas and predictions go up on a Web site for everyone to see, and many participants have bottles of wine wagered on the outcome.

  SEARCH stresses that this is a serious program to enhance scientific cooperation and not an Internet gaming site. Just in case anyone gets confused, they publish a big, bold “DISCLAIMER,” embarrassingly like those you see on dodgy stock market tip sheets, saying “This Sea Ice Outlook should not be considered as a formal forecast or prediction for Arctic sea ice extent…”

  In May 2008, nineteen groups sent in an “outlook.” Initial forecasts ranged from 1.1 to 2.2 million square miles (2.9 to 5.6 million square kilometers) for the area of ice that remained at the end of summer. Remember that the record one-month low after the great crash of 2007 was 1.6 million square miles (4.3 million square kilometers) and the average over the previous decade had been 2.6 million square miles (6.7 million square kilometers). The high forecasts came from “conservatives” who thought sea ice would recover toward its long-term trend, the years when it was just losing a steady 10 percent of its area per decade. The low numbers came from “catastrophists” who saw a second crash hurtling toward them. As it turned out, the final 2008 summer sea-ice minimum was 1.8 million square miles (4.7 million square kilometers), a small recovery from the great crash in 2007. Only one group came close with its forecast, but the median of all forecasts was even closer, suggesting that wisdom is best found in the crowd. The best part of the game is that it enables ordinary people to follow how scientists are developing their thinking about the ice and the logic behind their forecasts.

  We can’t really expect perfection, as we’d need to know the weather coming to the Arctic in advance plus have a very accurate map of that year’s starting conditions, which would include a map of the thickness of the ice that we can’t yet make. Given the uncertainties, scientists seem to be doing well. But new links and feedbacks are still being discovered.

  David Barber, a climatologist from the University of Manitoba, led a remarkable project which sent the Canadian icebreaker Amundsen out to the Beaufort Sea from September 2007 until August 2008, right though the winter. It spent much of its time off Cape Bathurst where a giant lead—a long, wide fracture—usually opens up in the ice. No one had ever spent the winter out in the Arctic on a research icebreaker before, so there were plenty of surprises.

  “We were expecting ice to form in October, but by the end of November we still had no ice,” says Barber. The first part of the explanation is familiar. In the long, hot summer of 2007 the open waters had soaked up heat and were just too warm to freeze quickly. “Climate stations on land were reporting air temperatures of-20°C, but over the oceans it was wide open and temperatures were much warmer, around-2°C,-3°C,” says Barber. Then came the surprise. Huge areas of warm air above the ocean fed storms that dumped snow onto the remaining ice. “A meter of fluffy white snow built up and insulated the multiyear pack ice from the cold air so it did not grow as much as it should have done,” explains Barber. That left thinner ice to face the summer melt.

  So here is yet another little loop to add to the thicket of interrelated changes we have seen in the Arctic: open water warms up in the sun, warmer water can alter storm tracks, storms drop more snow on ice floes that insulates them from the cold, the ice can’t grow, the ice melts away faster in summer leavin
g more open water, and so on. “Scientists have never had the ability to measure these kinds of processes in a detailed way before,” says Barber.

  When you start to see so many different interactions, often subtle and unexpected, affecting the Arctic ice, the big question that springs to mind is how many of them have a clear connection to the greenhouse gases that we have been pumping into the atmosphere and how many of them are “natural variation,” the usual ups and downs of the Arctic.

  To put the question more bluntly, do we have to accept that we are to blame for the loss of the sea ice, a question I’d been asking myself ever since I saw that first movie of the changing ice? That question quickly prompts another. Given the speed with that the Arctic is changing, how long will it be before we start to see summers that are free of ice? There are no more vital questions than these for the future of the Arctic seas, the creatures that live there, and the people who make their home around its shores. The only way to tackle them is by turning to computer models of the changing climate.

  Based on all the evidence we have, we can be very confident that the Arctic is warming as a result of the rising levels of greenhouse gases in the atmosphere. The most general test came recently when the Climate Research Group at the University of East Anglia looked at all the global climate models that had been used by the Intergovernmental Panel on Climate Change (IPCC), the international body that was set up to provide objective advice on climate change.8 “Everyone knows that temperatures in the Arctic are rising,” Philip Jones, the director of the group, explained to me in 2008, “but the Arctic region had not been looked at on its own until now in computer models of changing global climate.”