After the Ice

by Alun Anderson

  By that time, a bigger picture had emerged that confirmed the early work of the Royal Navy. In 1998, the U.S. Navy finally agreed to release a treasure trove of old data gathered in scores of submarine voyages all over the Arctic, some of it forty years old. Already in the early 1990s, following the collapse of the Soviet Union, U.S. Navy submarines had begun welcoming scientists aboard submarines on special expeditionary cruises in the Arctic. Still, it took an extra push from then U.S. Vice President Al Gore to obtain older data along with maps of submarine tracks. Even so, the data was deliberately made less precise before release and did not cover areas close to Russia.

  A team of scientists led by Drew Rothrock of the University of Washington in Seattle was the first to analyze the submarine data.3 They showed that the sea ice thickness in the 1990s had fallen by more than four feet, from the average of over ten feet recorded some twenty to forty years earlier. The volume of ice had fallen by around 40 percent, just as Wadhams had estimated. And the changes seemed to be happening everywhere that the submarines had been. Ice was not simply thinning in one area while thickening in another, which could be a result of winds shuffling ice around.

  “The submarine study that compared the measurements in the 1970s and 1980s with cruises in the 1990s was spectacular—they were the most significant and dramatic results we had,” says Christian Haas. Frustratingly, they are still nothing close to enough. They only provide a few data points on a graph. They can’t show what changes happened year by year or where in the Arctic they happened. What was really needed was accurate, Arctic-wide information on the changing ice thickness, preferably going back many decades to provide scientists with the clues they needed to solve the big question of why the ice is vanishing so rapidly.

  Without easy answers, scientists had to figure out some ingenious ways of reconstructing the history of the ice. One group, from the University of Colorado and NASA, led by Jim Maslanik and Chuck Fowler, came up with a cunning way of taking information from several different satellites, each built for a quite different purpose, and then combining that data with information from yet other experiments designed to track buoys drifting through the Arctic.4 They weren’t able to provide direct measurements of ice thickness, but they did come up with the next best thing: how the age of the ice had changed in different parts of the Arctic. Ice age is a reasonable stand-in for thickness because older ice that has survived many summers and passed through many winters is thick. New ice, naturally enough, is thin.

  I called up Jim Maslanik in his laboratory at the University of Colorado and asked him how and why they had reconstructed the past. “These different types of ice are pretty fascinating and I’d been interested in them for some time,” Maslanik explained. “But back in the 1990s, there hadn’t been the real imperative for studying them because we weren’t seeing the big changes in the Arctic sea ice then. With hindsight we should have picked up on it earlier. As time went on, we began to think, ‘Boy, it would really be good to have a way of measuring ice age.’ I wanted to come up with something that would not just say that this ice had survived one or two melt seasons, but whether it had survived three or four melt seasons. That was impossible from standard satellite approaches.”

  If ice weren’t constantly moving around the Arctic, finding the age of a particular patch of ice would have been pretty easy. You could just look at a series of satellite pictures of a particular area taken over the years and ask: “When was open water last seen here?” The year that spot froze over—the year the ice was “born”—would tell you how old the ice was now.

  Scientists had access to plenty of pictures of the Arctic taken from satellites in NASA’s Pathfinder series, which flew from the 1970s to 2007. Many pictures had been taken using infrared light—beyond the reddest part of the spectrum we can see—or microwave radiation, which is even further along the spectrum toward radio waves. At these wavelengths, ice and open water show up as different colors, so if ice stood still, some quick answers would have been within reach. But ice is constantly on the move.

  “We were kicking around some ideas about measuring ice age,” Maslanik continues. “Chuck [Fowler] said he knew of a way to measure ice movement, so maybe we could just try to see if we could estimate how long the ice survives by tracking its movement.” They took data from the tracks of buoys that had been left out on the ice. There were enough of these buoys scattered about the ice that they could divide the Arctic into a grid and then work out, from their tracks, how each of those grid areas had changed position and shape over the years.

  Combining charts of ice movement with the pictures of the Arctic ice taken from various satellites in some clever computer programs gave them the answer they were looking for. They could follow areas of ice as they moved around the Arctic and track which ice survived the succession of summer melts for one year, two years, three years and so on—and which did not.

  Out of this work sprang a map which showed the changing age of the Arctic’s sea ice, going back over several decades. “We were surprised at just how realistic the maps were. We’d combined all these different kinds of data and then put them together with the motion calculation. You know if we had sat down to write a proposal to ask for funds for this work we probably never would have got it,” says Maslanik. “We would have been told that too many different errors would add up and it wouldn’t work. Amazingly, it looked really good. So that gave us ice age.”

  That was a great first step, says Maslanik, but the really important thing people want to know is the thickness. He knew that first-year ice was pretty much between three and five feet thick. Multiyear ice, ice that survives one or more melt seasons, is typically between six and ten feet. But what about the ice that is five years old? Can you use an age estimate to provide a thickness estimate? Submarine and other ice-thickness data provided comparison points. It turned out that ice age gave a reasonable estimate of ice thickness, but as the ice grew older, the estimates grew less accurate.

  Such, then, was the long and winding road by which a dozen or so scientists were able to create maps of the changing age of the ice in different parts of the Arctic and, along with it, a good idea of changing ice thickness.

  The maps, published at the end of 2007, provide a grim picture of the rapid disappearance of the older, thicker ice that fits well with the broad trends seen from under the ice by submarines. In spring 1985, areas of older ice—where at least half the ice was five or more years old—covered 2.2 million square miles, or more than 36 percent of the Arctic ice. In spring 2007, that area of older ice had fallen by more than half to just 1 million square miles, 17 percent of the 2007 total area. The percentage of multiyear ice five or more years old fell from 31 percent to 10 percent, while that seven or more years old fell from 21 percent to 5 percent. The Arctic ice was clearly growing ever younger, thinner, and more vulnerable.

  That data is for ice age, not true ice thickness, but the overall picture of vanishing older, thicker ice fits the all too few direct observations that have been made of ice thickness across the Arctic. Data from Haas’ EM-bird, for example, shows that around the North Pole, ice that had a modal thickness of a little over 8 feet in 2001 had thinned to 6.5 feet in 2004 and to 3 feet in 2007.5

  Scientists are very far from having a complete, detailed picture of the changing ice thickness that they need. If Cryosat 2—being built to replace the satellite designed to measure ice thickness that crashed near the North Pole—is launched successfully, that will be a big help. Still, with the evidence that we have, we can be sure that the ice is growing younger and thinner, and it is likely that ice volume is falling even faster than ice area is shrinking.

  That takes us back to our big questions: what is making the ice thinner and what connection does it have to global warming? We have already explored one answer to the first part of that question. Changes in the pattern of winds can change the flow of ice around and across the Arctic and push ice out of the Arctic. If old ice were pushed out of the Arctic, t
he thickness of what remained would be smaller. What else is important?

  This is where we start to run into differences of viewpoint. Some scientists stress that the ice is growing thinner because of changes up top, in the atmosphere. Average air temperatures in the Arctic have risen by 2.5°C (4.5°F) over the last twenty years, and summer sunshine can quickly warm the surface of the summer Arctic seas in which the ice sits and eat away at it, while winds shuffle it around the Arctic. Others stress that the ice is also melting from below, or its winter growth is being hit, as a result of heat coming up from deeper in the seas. Many believe that the ice is simultaneously under attack from above and below but are not sure of their relative importance and point out that the answer won’t be the same everywhere in the Arctic.

  At this point I have to smile. Scientists are trained to look objectively at all the evidence. But, being human, they are most excited about the particular topic they are studying. Ask an atmospheric scientist if the Arctic is thinning because of changes to the ocean or the atmosphere and you’ll get a quick reply: “atmosphere.” Changing weather, winds, air temperature, and sunshine are where we will find an explanation. Those who think the ocean has a big role turn out to be mostly oceanographers. They’ll be looking at temperature changes in the sea, warm currents, and the layering of the ocean below the ice for an impact on the way ice grows.

  So if you talk to an atmospheric scientist and then to an oceanographer (or vice versa), you can end up confused. Talk to a roomful of them, and your mind will be in a complete spin. So I’ll take them one at a time, and start by plunging back under the ice, because that turns out to be a more controversial place to be. Straight away we encounter something very strange (at least to me): down there beneath the sea surface, the water is quite warm.

  One of the very oddest things I learned about the Arctic is that its ice is sitting on top of seas containing more than enough heat to melt all the ice away many times over. On the Atlantic side, from 650 feet or so on down, there is a layer of water that has come in from the much warmer oceans to the south. That water doesn’t just pour into the Arctic as though you were filling a bathtub, but flows in where it most easily can. The Atlantic water creeps in through the Barents Sea and the Fram Strait and travels around the coast of Siberia to the Laptev Sea, where some is pushed by the huge Lomonosov Ridge toward the Amundsen Basin and some travels onward, deeper into the Arctic. On the Pacific side, warm water from the Bering Sea slips in across the shallow Bering Strait and slides under the cold surface waters. What depth these waters end up at is mostly determined by their density (unless something is mixing them) and that depends on both their salinity and their temperature. Other things being equal, salty water is denser than freshwater and cold seawater is denser than warm seawater.

  These deep waters are growing warmer. “From an Arctic perspective, it is an invasion of warm water coming from somewhere else. But really it’s a warming of the world ocean. The warming of some other part of the ocean comes to the Arctic Ocean eventually. So that’s going to continue just as global warming continues,” explains Mike Steele, an oceanographer at the Polar Science Center at the University of Washington in Seattle with a long interest in the layering of the Arctic seas.

  If global warming means warmer oceans and those warmer oceans invade the Arctic, then we surely have an explanation of why the ice is thinning. That at least is what I thought when I first heard about the warm waters beneath the ice, but unfortunately it is not so simple.

  In order to find out why, we will start on the Atlantic side of the Arctic, as east and west are quite different. Down below the Arctic ice, there is an enormous reservoir of warm Atlantic water. But its heat is trapped. Sandwiched between the deeper Atlantic water and the top layer of the sea is a layer of cold water, fresher at the top and saltier below, that insulates the surface from the depths and stops its heat from escaping. It is called the cold halocline layer (the word “halocline” tells us that it has a gradient of saltiness).

  When I expressed my astonishment that there is just a thin layer of icy cold water keeping the ice from melting away completely, Steele, who has been studying the halocline layer for more than twenty years, chuckled and said, “Yes, it is kind of cool.”

  At the very top of the Arctic Ocean is a layer of relatively fresh water, some 150 to 300 feet deep. This is where ice forms and, not surprisingly, the water is usually at a temperature close to freezing. This topmost, cold, fresh layer (the “surface-mixed layer” as scientists call it) is fed by the great Siberian rivers pouring cold freshwater onto the surface of the Arctic, by rain, and by the melting of ice in summer. As the water is fresh, it is light and floats on top of denser, saltier waters beneath. Deep down below are the warm, salty Atlantic waters. In between is the critical insulating cold halocline layer which is fed by the formation of sea ice. When seawater freezes in winter, salt is squeezed out of the growing ice crystals, leaving behind freshwater ice and cold, dense, briny water that spreads out below the lighter, surface-mixed layer. Polynyas, areas of the Arctic where the seas are regularly open in winter, help maintain this cold halocline layer. Within polynyas, newly frozen ice is constantly pushed away by winds and currents, exposing yet more open water to freeze. These “ice factories” and the brine they reject help fuel the cold halocline layer that spreads across the Arctic.

  That creates an astonishing situation. Jean-Claude Gascard, leader of Europe’s DAMOCLES project of Arctic observation put it like this: “If this cold halocline, which has been formed in winter, disappeared, there is nothing to stop ice from melting. All the water underneath, all the air masses above are well above the melting point and ice will melt very rapidly.”

  Where the cold halocline is thin, heat will more easily escape from the warm Atlantic layer. Peter Wadhams is sure that the warming of the Atlantic layer below the ice has been important in the area he looked at, where the halocline layer is perhaps not as thick as in the center of the Arctic. “The most rapid thinning of the ice seemed to have happened in the mid-1980s to mid-1990s, which was when we noticed a distinct warming of about a degree in the core of the Atlantic layer of the ice. That means more heat was being pumped in from the Gulf Stream—and the extra degree is enough to increase the heat flux from below. It was probably one of the leading causes of the thinning of the late 1980s and early 1990s,” he says.

  Closer to the North Pole, where Christian Haas took his recent measurements of ice thickness, ocean heat doesn’t seem to be coming through. “In places where the top layer is only relatively thin, ocean heat may come through, however in most regions of the Arctic, once this water has submerged below this mixed layer, it doesn’t have any effect on the ice anymore,” he says.

  Across the whole Atlantic side of the Arctic there is not enough data to be sure if the warm, deep waters are having a growing impact. Nor do we know what might happen in the future. “Thanks to the halocline this ice is not melting now,” says Gascard. “The problem is what’s going to happen with so much open ocean and with the halocline exposed to the storms.” Storms may mix the surface waters, erode the halocline layer, and reinforce the destruction of the ice. A catastrophe may be waiting.

  Over on the other side of the Arctic, not far from Alaska, where warm water enters the Arctic from the Pacific through the shallow Bering Strait, there is far more evidence that heat is already leaking from below and destroying the ice above. Here the ocean may be playing a much bigger role. This is the side of the Arctic that has seen the really massive ice losses.

  Take a look again at the map of the September sea-ice minimum for 2008 and compare it to the map of the ice at the end of March when the winter ice is at its maximum. The ice hasn’t shrunk back steadily from the Atlantic and Pacific sides of the Arctic toward the North Pole. Rather, it looks as though something came out of the Pacific, passed through the Bering Strait, and gobbled up all the ice within reach.

  Koji Shimada, the Japanese researcher who showed me that sped-u
p horror movie of the Arctic’s whirling ice, thinks he knows what that is, although his ideas are controversial. Under the ocean in the area where so much of the ice has gone missing, there is a large sea mount called the Northwind Ridge which is linked to the Chukchi Plateau, part of the continental shelf that fringes Alaska and Russia. In the past, Shimada explained, few ships had explored the area. The most popular route for research cruises had been between Barrow, on the northwestern tip of Alaska, and the North Pole. But more recently, trips to the area found that just below the surface there was a “very warm strong current along the Northwind Ridge, a kind of Arctic Gulf Stream.”

  But wasn’t the cold halocline trapping heat below the surface here, too? Not really, explains Shimada.6 On the other side of the Arctic, the Atlantic water circulates at a depth of between 1,000 and 1,300 feet, much deeper than the Pacific water; it is denser, and it is not easy for it to reach to the surface. Over on the Pacific side, the water is not strongly layered. Shimada showed me a couple of profiles of the sea around the Northwind Ridge, with bright colors indicating the temperatures at different depths. The warmth of the ocean comes right up close to the surface. “The surface water density is almost identical to the warm water. That is the key point,” he says. The two waters can mix and there is no real barrier between them.

  That heat has some subtle and powerful effects. “Most people who think about the reduction of the sea ice in the Arctic Ocean just think about the ice melting,” he said. “It is more important to think about it as the imbalance between the ice formation that happens in winter and the ice melting that happens in summer. This is the essential issue. What happens in winter is very important.” So, I needed to think beyond “faster ice melt” in the summer and include “slower ice growth” in the winter. Either way, you end up with thinner ice.