This site uses cookies. By continuing to use this site you agree to our use of cookies. To find out more, see our Privacy and Cookies policy.
Skip to the content

IOP A community website from IOP Publishing

Powered by Movable Type 4.34-en

In from the cold: October 2009 Archives

The Atlantic keeps cropping up when we try to understand why glaciers change. If you look at the right kind of map, you can see that the Atlantic, including the Arctic, is an enormously long inlet in the shore of the world ocean. (The Bering Strait, between Alaska and Siberia, is too shallow to make a difference.)

The ocean is hot near the equator and not so hot near the poles. The heat flows down the temperature gradient, which drives the ocean currents. The water has to go somewhere once it has got where it is going and has surrendered the heat to the atmosphere. So it sinks. The sinking has a natural explanation: now that the water is colder, it is also denser.

At this point, the first two of several intertwined complications alter the picture. First, as it surrenders heat, the north Atlantic also surrenders water vapour. The warmer the water, the faster it evaporates. Second, during evaporation the salt stays behind, making the ocean water denser than it was to begin with, and that makes it yet more likely to sink.

The water that sinks gets back where it started from by flowing southwards at depth. Eventually it finds its way out of the Atlantic inlet and wells up, or is dragged up by the wind, in broad regions of the low-latitude ocean, where it is reheated by the Sun and thus begins the cycle over again. This is the essence of the meridional overturning circulation, or MOC. The Atlantic part of this circulation, predictably abbreviated AMOC, is a sensitive part of the machine because its northern end is where most of the Northern Hemisphere overturning happens. There is nothing comparable in the north Pacific.

The water vapour released from the north Atlantic adds greatly to the poleward deliveries of vapour through the atmosphere, and helps to make the northern shores of the Atlantic snowy. The glaciers around the north Atlantic today, and the much bigger ice sheets we had during the ice ages, owe a lot to the AMOC.

But now we come up against the third of the intertwined complications. After the water vapour has condensed and fallen on the bordering landmasses, it flows back sooner or later to the ocean, where it dilutes the salt that helps the AMOC through the crucial sinking part of its cycle. In principle, a large enough return flow of fresh water from rivers and glaciers could reduce the density of the surface waters sufficiently to stop them from sinking, in which case the whole AMOC would stop.

And now, enter the fourth of the intertwined complications: in one word, us. If we heat the whole system by enough to shrink the Greenland Ice Sheet significantly, flooding the north Atlantic with fresh water, we raise the prospect of just such a switching-off of the AMOC.

All the climate models suggest that, if the AMOC collapsed, the northward heat transfer would also be greatly reduced and the shores of the north Atlantic would suffer cooling. But fears of a new ice age being triggered by a collapse of the AMOC, itself triggered by a collapse of the Greenland Ice Sheet, are not realistic.

In the first place, these collapses would happen in a context of global warming, and again the climate-model evidence shows that they would not suffice for the job. Secondly, we have lots of palaeoclimatic evidence for abrupt changes in the AMOC, which are leading candidates to explain Dansgaard-Oeschger transitions during the last ice age, and the cold snap 8,200 years ago. They didn't last all that long, and they were all reversible. Thirdly, models of ancient climates suggest persuasively that the AMOC is not implicated as a mechanism for starting ice ages.

And finally, the models agree that, without actually collapsing, the AMOC is nevertheless very likely to weaken over the next century. Even decanting the Greenland Ice Sheet into the ocean would not switch it off. But several metres of sea level rise, and a weaker AMOC in a warmer world, are enormous problems in themselves. That they are not harbingers of a colder world is not a good reason for relaxing.

In glacier monitoring, one of the things we worry about is undersampling. The measurements are sparse, and we have to interpolate, that is, make plausible guesses about the glaciers we can't measure. Gaps in coverage mean that there is always a chance that new measurements in remote areas will change the picture. One of these areas is the Subantarctic islands, scattered across the Southern Ocean and holding about 8,000 km2 of glacier ice in all. Our knowledge of this ice has been fragmentary until recently. Could the Subantarctic be an exception to the global rule of glacier shrinkage?

The knowledge base is beginning to improve, and we can now say that the answer is "No". For example, with Étienne Berthier, of the Laboratoire d'Études en Géophysique et Océanographie Spatiales in Toulouse, I am writing a chapter on the Subantarctic for a book about GLIMS, the Global Land Ice Measurements from Space initiative. Étienne kindly sent me an April 2009 ASTER satellite image of the west coast of Kerguelen, in the southern Indian Ocean.

Glaciers in Kerguelen, 1963-2009 (1963 outline in yellow)
Glaciers in Kerguelen, 1963-2009 (1963 outline in yellow)

The small protruding glacier tongue in the lower part of the picture belongs to Glacier Pierre Curie, which now ends a kilometre from the sea but in 1963 had a calving front about 600 m wide. The two stubby tongues in the upper part are Glacier Pasteur, whose calving front was 1700 m wide in 1963 but is now only barely in contact with tidewater. In another few years, it will have retreated away from the shoreline of Anse des Glaçons (the cove of ice floes, a place name which will provoke nostalgia one day).

The retreat of these two adjacent outlets of the Cook Ice Cap doesn't count, nowadays, as startling news. Berthier and his co-authors recently reported that Cook Ice Cap shrank at 2.4 km2/yr, half a percent per year, between 1963 and 2001. At this link you can watch an animation of the shrinkage of Glacier Ampère, on the opposite side of the ice cap from Pasteur and Pierre Curie. But if you picked any two neighbouring glaciers almost anywhere in the world, the odds are that they would have shrunk at something like that rate, or perhaps a bit less. So now we know that Kerguelen was not one of the out-of-the-way places where a surprise was awaiting us. The Kerguelen glaciers even follow the widely-observed tendency of accelerating shrinkage (that is, faster recently than earlier).

Another out-of-the-way place about which we now know a lot more is Heard Island, in the Indian Ocean southeast of Kerguelen at 53° South. It has tidewater glaciers mainly because it rises to 2,755 m above sea level. My other co-author, Shavawn Donoghue of the University of Tasmania in Hobart, finds that Heard Island's 30 glaciers are dwindling just as are those of Kerguelen. Six have parted company with the sea during the decades since the first air photos in 1947, leaving a dozen still delivering icebergs to the ocean. Gotley Glacier, which drains the summit crater of Big Ben, is still standing in the sea as it has done for as far back as we have information.

Calving glaciers are more challenging than ordinary ones when it comes to documenting change. Those that manage to advance far down a fiord can misbehave spectacularly. More often, however, as with Gotley and its 11 neighbours, the icebergs break off as soon as the ice reaches sea level, so the calving front doesn't change much.

Change in a previously unknown region that turns out to be globally typical – "Subantarctic Glaciers Not Surprising" – is difficult to sell as a motive for political action. What we are after here, apart from a conversation piece for your next cocktail party or trivia game, is something that will inject the necessary urgency into the deliberations of the politicians and policymakers. They will assemble in Copenhagen this December for the most important negotiating session in the history of the human race, the fifteenth Conference of the Parties to the UN Framework Convention on Climate Change. Faced with the evidence, they seem to have got the scientific message, but it hasn't really clicked yet. Graphs that fall off the bottom of the page haven't done the trick. Neither Gotley's continued stillstand nor Pasteur's impending loss of tidewater status are likely to make the communications breakthrough, but you never know.

Dynamic thinning is even more exciting than basal lubrication, but recent work by Hamish Pritchard and colleagues shows up the difference. For basal lubrication of the flow of glacier ice, just water will do, but it is now clearer than ever that for dynamic thinning what you need is warm water. Or rather, and this is not the same thing, if you supply warm water what you get is dynamic thinning.

Dynamic thinning is thinning over and above what can be expected from an imbalance between the ability of the climate to generate meltwater and the ability of the glacier to replenish the stock by flow. There is a simple reason for getting excited when we observe dynamic thinning. It means that the ice must be moving faster, and shedding more mass at its terminus, than it did before the dynamic thinning began.

Most glaciers are thinning at present. Strictly speaking, we mean that their surfaces are getting lower, but we can correct for the possibility that part of that is due to change of the bed elevation. Measuring the change of surface elevation is within the grasp of several methods. In particular, the ICESat laser altimeter on which the Pritchard work relies can give very accurate estimates.

In Greenland the most thinning you can expect from the climate – snowfall minus melting – is a few m/yr. In Antarctica, less than that is usual, and there are plenty of outlet glaciers down there where even today there is no melting even at sea level, and all of the loss is by discharge across the grounding line. Dynamic thinning, and the implied faster flow and faster discharge, are signs of trouble.

What Pritchard and colleagues have shown is that dynamic thinning was widespread around the margins of Greenland and Antarctica during 2003–2007, and is on the increase. Both of these findings are scary. There ought not to be any dynamic thinning in a well-regulated world. Perhaps their most telling finding is that there is a definite difference between fast-flowing tidewater glaciers and slower parts of the ice-sheet margins.

Roughly, the slow-flowing margins are showing just run-of-the-mill accelerating loss with no evidence for dynamic thinning. We are learning rapidly about basal lubrication as one of the reasons for the run-of-the-mill acceleration. But if it were the main or only reason then there would be no difference between the land-terminating and the tidewater margins. In fact, all hell is busting loose on the tidewater glaciers. Pritchard and colleagues have now documented dynamic thinning in the north of Greenland as well as the south, and in the east of Antarctica as well as the west – but only on tidewater outlets, and only on fast-flowing tidewater outlets.

These results point squarely at the ocean as the culprit. The better the ocean is at melting the base of a fringing ice shelf, or at sapping a grounded calving front, the faster does the glacier go in its efforts to maintain the supply of ice, and the greater the resulting dynamic thinning inland. At this point, our explanations of what is happening begin to get rather hand-waving. Measuring the temperature of the ocean near to tidewater glaciers is an extreme challenge, and we know very little about whether it is changing. But there are indeed some signs that warmer water is getting at the ice.

For example, David Holland and colleagues have shown dramatic maps, based on measurements by fishing boats, of the arrival of warm water in Jakobshavn Fiord in western Greenland. The warm Irminger Current, deriving from the Gulf Stream, curls clockwise around southern Greenland. Between 1996 and 1997, it flooded Jakobshavn Fiord with water 2–4 °C warmer than what was there before. In 1997, the floating tongue of Jakobshavn Glacier, the largest outlet of the Greenland Ice Sheet, began to disintegrate, and simultaneously the dynamic thinning of the glacier began. By 2007 it was 30 m thinner 70 km inland. At 15 km inland, it was more than 200 m thinner. It is still thinning, and nobody can tell when it will stop thinning.

Not all of the dots connecting global warming to the dynamic thinning of Jakobshavn Glacier have yet been joined up, but I don't know anybody who is betting on this being the wrong explanation, or, after the demonstration by Pritchard and colleagues, on Jakobshavn being an isolated example. This makes the world look even more complicated, and for the moment at least it makes predicting the future contribution of glaciers to sea-level rise even harder.

During my first ever field season, I studied a small drainage basin on Devon Island in the High Arctic of Canada. There was a glacieret, a tiny glacier, in the upper part of the basin, and it was mostly pink. As I had been taught that glaciers are whitish or brownish, depending on how much sediment they contain, this pinkness surprised me.

Later I learned that coloured ice is not all that unusual. Glacier ice can be intensely blue if it is free of air bubbles, which are the source of the whiteness. It can be black, or at least look black, if it is floating on water and is transparent. But pink? In northwest Greenland there is a 70-km stretch of the margin of the ice sheet called the Crimson Cliffs. It turns out that the pink colour comes from the carotene manufactured by bacteria that contrive to get a living from the glacier surface. Relatives of this bacterial carotene explain the pinkness of flamingos (they will insist on eating shrimps, which eat the pink bacteria), and the carrotiness of carrots.

Biologists are a bit like glaciologists in that they are willing to study almost anything, so bacteria on glaciers are not new to science. But are they any more than an arresting curiosity?

In recent years, microbiologists have become quite excited about evidence for a so-called deep biosphere. The distribution and abundance of hydrocarbon molecules in deep environments suggest that there must be organisms manufacturing some of the molecules. These deep environments include the beds of glaciers. From the Arctic, the Alps and elsewhere, strong circumstantial evidence has accumulated for ecosystems consisting of distinctive subglacial microbes. Last year D'Elia and colleagues showed photomicrographs of bacteria and possible fungi from ice that had accreted from the water of Subglacial Lake Vostok in Antarctica. The microbiologists seem to be satisfied that they are not looking at samples contaminated by near-surface organisms.

Apart from provoking us to rethink the meaning of "life", subglacial microbes have implications. Wadham and colleagues explore the question of what they might have done to the climate if they were active beneath the ice sheets of the last ice age. There was plenty for them to eat, in the form of overridden rotting vegetation, which some of them would have converted to methane. When the ice sheets waned, the methane, a greenhouse gas, could have had substantial climatic impact when released to the atmosphere. Apparently the release would have had to be episodic to have made a big difference. This is pure, though constrained, conjecture – but what fascinating conjecture it is.

There is a potentially enormous payoff if we can develop an understanding of how organisms can thrive at the beds of glaciers. They may help to stretch the envelope of hospitability yet further, because the most impressive glacier in the known universe is one that is not on the Earth's surface at all.

Europa, one of the Galilean satellites of Jupiter, has an outer shell consisting mainly of water ice. We cannot be sure of the thickness of the shell, but it is probably a few tens of kilometres and may be as little as just a few kilometres. The most interesting things about this lithospheric glacier may be that, firstly, it is undoubtedly floating on an ocean of what is almost certainly liquid water, and secondly, that there is compelling evidence of resurfacing. That is, images of the surface of Europa show features that make sense only if the underlying ocean has on occasion managed to rupture its ice cover and spill out onto the surface of the satellite. So Subglacial Lake Vostok, and the beds of glaciers generally, are intensely interesting from the standpoint of the search for extraterrestrial life.

Remember that oxygen, a deadly poison, is irrelevant, and the one so-far-universal common denominator of known life-hosting environments is liquid water. Ice will not do, and nor will steam. The beds of glaciers have what is needed, and they host life. The first of these two assertions appears to be as true on Europa as it is on Earth.