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 2010 Archives

Dirt on glaciers

| | Comments (6) | TrackBacks (0)

Dirt is more or less ubiquitous on glaciers. Even when there isn't much, it makes the ice darker, a climatically important fact that can also be put to intriguing uses. Sometimes the dirt gives us striking visual effects in the form of multiple medial moraines. When there is enough dirt to cover the surface nearly completely, we call the glacier a debris-covered glacier.

The glacier isn't necessarily buried entirely. In the accumulation zone, fresh snow will mask older debris, and usually not all of the ablation zone, where all of the snow melts, is debris-covered. Indeed the debris typically covers only a few to perhaps 20 percent of the surface, at the lowest elevations. But this part is critical, because it is where we expect to observe most of the mass loss that is supposed to balance the mass gain in the accumulation zone.

Debris-covered glaciers are a nuisance from a number of standpoints. Probably the biggest nuisance is that we don't know how to judge the effect of the debris on the mass balance. Conventional wisdom has it that thin debris increases, and thick debris reduces, the melting rate of the underlying ice. The debris, being darker than the ice, absorbs more of the incoming radiation, but the more debris there is the less radiative heat reaches the ice.

In a recent laboratory study Natalya Reznichenko and co-authors reinforced the conventional wisdom. In an interesting twist, they showed that the daily cycle of radiation makes a big difference. If you irradiate a sample of debris-covered ice continuously, then after a delay of some hours, increasing with the thickness of the debris, you get the same rate of meltwater production as if there were no debris at all. But if you cycle your lamps with a 12-hour period, to mimic day and night, the underlying ice never gets a steady input of heat. The night undoes much of the work accomplished during the day.

The melt rate is slower beneath debris more than 50 mm thick, about half a handsbreadth, and faster beneath thinner debris. This is the "critical thickness", at which the debris has no net impact on the melting rate. In an elegant graph summarizing the field measurements, the Reznichenko study shows that the critical thickness varies with altitude or equivalent latitude. On higher-latitude glaciers, or equivalently at lower altitude, the critical thickness can be as low as 20 mm, but it can exceed 100 mm at low latitudes or high altitudes.

So the real world is a bit more complicated than the laboratory when it comes to the effect of real debris on real glaciers. What about real-world melting rates? There are distinct signs that real debris-covered glaciers are even harder to understand than ideal ones.

For example, Akiko Sakai and co-authors pointed out that on debris-covered glaciers in Nepal there are lots of small meltwater ponds. One such pond absorbed seven times more heat than the debris cover as a whole, accelerating the melt rate and producing a knock-on effect because the departing warm meltwater enlarged its own conduit, a phenomenon known as "internal ablation". And on a debris-covered glacier in the Tien Shan of central Asia, Han Haidong and co-authors showed that exposed ice cliffs, accounting for only 1% of the debris-covered area, produced about 7% of all the meltwater generated across the debris-covered area.

It would be nice if we knew the fractional debris cover of every glacier, and even nicer if we could say by how much the mass balance of each glacier is altered by its debris cover. This is a pipe dream for the moment, but the careful small-scale measurements on debris-covered ice seem to suggest that ignoring the debris, as we have to do now when estimating mass balance on regional and larger scales, could overestimate the mass loss quite seriously.

The trouble is that although we have very few measurements of the whole-glacier mass balance of debris-covered glaciers, they seem not to be consistent with the detailed laboratory and field studies. One of the most careful regional-scale measurements, by Etienne Berthier and co-authors, found that Bara Shigri Glacier, in the Lahul region of the western Himalaya, thinned by —1.3 m/yr over five years, a rate significantly faster than the —0.8 m/yr determined by photogrammetry for all the glaciers in the region.

"Bara Shigri" means "great debris-covered glacier" in Hindi. Clearly we don't know enough about the behaviour of debris-covered glaciers, great or small, and as there are lots of them, and lots of people depend on them, they need continued study.

You can find all sorts of things in glaciers if you look hard enough. Among the oddities that come to mind are volcanic sulphur, soot, and bacteria and fungi.

But now Andrei Kurbatov and co-authors, writing in the Journal of Glaciology about fieldwork on the western margin of the Greenland Ice Sheet, have found something really surprising: diamonds. Don't get too excited. If you look in just the right place you can expect to find trillions of them per litre of melted ice, but these are nanodiamonds, the biggest only a few hundred billionths of a metre across. There is no danger of prices collapsing in the international diamond market.

However there is definitely a likelihood of a diamond rush spearheaded by scientists. The stimulus for this work was the discovery of nanodiamonds in ordinary sediments from several sites across North America. At all of the sites, much of the diamond is actually lonsdaleite, and there are other indications that point to the material being non-terrestrial. Lonsdaleite is elemental carbon that has crystallized in the hexagonal system, so it is a polymorph of the more familiar diamond belonging to the cubic system. Cubic diamond forms at temperatures and pressures appropriate to depths greater than about 150 km beneath the Earth's surface. To make lonsdaleite it appears that you need much greater temperatures and pressures even than that. At any rate, it is known only from meteorites and impact craters. We conclude that either it arrived with the meteorite or it formed during the impact.

The next exciting thing about these non-terrestrial diamonds is their age. They are found exactly at the base of the Younger Dryas cold snap, dating to about 11,000 BC. You could not ask for a sharper spike in abundance than the one shown in the Kurbatov paper, and it matches the evidence from elsewhere perfectly.

The first synthesis of this evidence showed that there are non-terrestrial "event markers" all across North America at the base of the Younger Dryas. It was a bold, if partly conjectural, synthesis, linking the impact not just to the cold snap but to the extinction of the mammoths, the disappearance of the palaeoamerican Clovis culture and the formation of the Carolina Bays.

The Carolina Bays can be seen in the atlas as the multiple arcs that form the coastline of the two Carolinas, but inland from the coast there are also numerous lakes of elliptical outline. They might be just quirks of Nature, but they would also be consistent with the putative Younger-Dryas impact having been in fact an airburst, followed by the impact of multiple smaller fragments.

We are now unambiguously in the realm of conjecture, but the Carolina Bays have been a geographical puzzle for centuries. Perhaps they are about to turn out to be not just a puzzle, as happened with the jigsaw fit of western Africa and eastern South America. Whatever their status, we can expect an energetic search over the next few years for the locality of the impact or airburst at the base of the Younger Dryas. We can also expect energetic discussions about its efficiency as a trigger for cooling.

The Greenland nanodiamonds are thus a small part of what is beginning to look like a much bigger picture, but they also represent a glaciological tour de force. I said that the Kurbatov spike was found "exactly" at the base of the Younger Dryas. So it was, but not in an ice core, as you might have guessed. The authors went to the ice exposed in the ablation zone about 1 km in from the margin of the ice sheet. All of it must have travelled some hundreds of kilometres from where it fell as snow in the ice-sheet interior. The text is rather coy here: "One of the authors (Jorgen Steffensen) ... identified a candidate for the Younger-Dryas-age section based on visual inspection of dust stratigraphy."

The atmosphere is slightly dustier when it is colder, and the dust makes ice that accumulated during cold episodes greyer. The nanodiamonds were found at the base of a band of greyish ice bounded above and below by whiter ice. There is therefore a sense in which the "exact" location of the base of the Younger Dryas has only been pinpointed circumstantially. But I doubt that there will be much questioning of the identification, and the success of the search is not less astonishing and gratifying for being due to use of the human eyeball as a search tool.

Chains of reasoning can be quite long, and quite tortuous, but they can join up the most surprising places. Consider beryllium-ten.

Almost all of the Earth's beryllium is beryllium-nine, the isotope symbolized as 9Be and defined by having four protons and five neutrons in its nucleus. We also have a small stock of 10Be, which has an additional sixth neutron. Most of the 10Be is created in the atmosphere when incoming cosmic rays collide with gas molecules: it is a cosmogenic nuclide.

Cosmic rays are not rays but particles, mostly protons, that originate outside the solar system. Some are energetic enough to destroy the molecules with which they collide. For example, not only may a molecule of nitrogen be split into its two constituent atoms, but one of those atoms, a 14N (seven protons and seven neutrons), may be split into a helium (4He) and a 10Be.

10Be dissolves in rainwater, which is weakly acidic. When the rainwater falls on land, it becomes more alkaline by reacting with the surface minerals, which induces the 10Be to precipitate out. Unless it is carried away by running water or the wind, it accumulates.

The point of the story so far is simple: the longer a patch of surface has been accumulating the products of cosmic-ray collisions, the greater its stock of 10Be. If it remains in place, as is likely on the surfaces of large boulders, we can count the 10Be atoms, correct for the slow radioactive loss (10Be has a half life of 1.36 million years), and obtain the exposure age of the surface by a relatively straightforward calculation — as long as we know the rate of arrival of the cosmic rays.

But here we come to a tangle in the chain. The cosmic-ray flux is not constant. The list of corrections that have to be made is quite long, allowing for factors such as the varying strengths of the solar wind and the terrestrial magnetic field, the altitude and latitude of the exposed boulder, and the extent to which it is truly "exposed" and not shielded by the surrounding terrain, nearby obstacles, seasonal snow and so on.

The allure of that exposure age, however, has stimulated intense efforts over the past couple of decades. We have developed a good understanding of many of the required corrections, and, like the atoms on the boulders, reliable 10Be exposure ages are now accumulating in the literature.

Writing in Nature, Michael Kaplan and co-authors offer a new collection of 10Be ages from the terminal moraines of a glacier that once occupied a cirque in the Southern Alps of New Zealand. Three of the 37 ages are oddballs, but the remainder all cluster nicely, in groups from different moraines and morainic ridges.

The outermost moraine, about 2 km from the cirque headwall, has boulders that were first exposed to the atmosphere just before 11,000 BC. In sequence, the moraines nearer the headwall have ages of about 10,700 BC, 10,100 BC, 10,200 BC and finally, only a couple of hundred metres from the headwall, 9,500 BC. All of these ages are uncertain by 400-600 years.

The payoff of these observations is in the dates, which span very neatly the cold snap of the Younger Dryas. While abundant evidence for cooling was piling up in various northern palaeoclimatic archives, this little New Zealand glacier was dwindling into nothingness.

Kaplan and co-authors have thus nailed down, much more firmly than before, the conclusion that the hemispheres were out of sync at the end of the last ice age.

The atmospheric concentration of carbon dioxide increased during the Younger Dryas, ruling out a reduced greenhouse effect as an explanation for the northern cooling. The generally agreed explanation is that the north Atlantic received a rapid influx of buoyant fresh water from North America. This reduced the overturning of ocean water, and damped down the meridional circulation. But it also pushed the climate belts southwards, shifting the southern-hemisphere westerlies to higher southern latitudes at which they were better able to provoke oceanic upwelling. The resulting enhanced outgassing of deep-ocean carbon explains the increased atmospheric CO2, and the 10Be atoms on the boulders show that the Younger Dryas was a time of net warming at least as far north as New Zealand.

So beryllium helps us to understand the hemispheric asymmetry of glacial climate. Long and tangled the chain of reasoning may be, but it does illustrate how complexity can be unravelled given doggedness and ingenuity.

Everyone knows something about Ötzi, the Iceman of the Ötztal valley in the Tyrol of Austria. For example, many know that he was found, slightly embarrassingly for Austria, just inside South Tyrol, which is on the Italian side of the frontier. He is now at rest in the South Tyrol Museum of Archaeology in Bolzano.

Schnidi, by contrast, is less well known. There are reasons. First and foremost, he doesn't exist. He, or quite possibly she, is a collage of human detritus spread over the northern approach to the Schnidejoch, a 2,756 m pass in the Bernese Alps in southwestern Switzerland. More remarkably, Schnidi is spread over 6,500 years of Alpine prehistory. Like Ötzi, though, he has a lot to tell us, and a lot to ask us.

The archaeological finds at the Schnidejoch are documented by Martin Grosjean and co-authors in the Journal of Quaternary Science. They were exposed by the recession of a small ice patch, recently detached from the larger Tungel Glacier, during the record-breaking hot summer of 2003.

This is the second remarkable thing about Schnidi. Among the clothes he discarded were perishable goatskin leggings and shoes, from as long ago as 4,500 BC according to new results announced in 2008. Apart from making the early part of Schnidi a good deal older than Ötzi (about 3,300 BC), this means that 4,500 BC was the last time it was as warm in the Bernese Alps as it is now. The leggings must have been preserved beneath the ice since then.

The Schnidejoch is not a particularly hard climb, but a kilometre or two downvalley a moderate advance of Tungel Glacier from its modern extent would close off the valley, making the route difficult if not positively impassable. This is the simplest explanation for the next remarkable thing about Schnidi. He clusters in time. There is late Neolithic clothing and hunting gear from 2,950 to 2,500 BC; arrows, pins and other material from the early Bronze Age (2,150 to 1,700 BC); shoe nails, coins and a woollen tunic from Roman times (the first century BC to the second century AD); and a few items from mediaeval times.

These intervals coincide rather well with nearby evidence for warm periods, but they are also complementary because Schnidejoch is much higher than the other sources of information, and it is a "binary and non-continuous archive" — the pass was either open or closed.

Ötzi and Schnidi raise all sorts of questions, some sobering and some frivolous. Why do we westerners see Ötzi as someone who can tell us things, while aboriginal Americans see his counterparts on their continent as in need of re-burial, to be left in peace? All I can offer is the reflection that it is a pity we can't tell things to Ötzi, and the thought that if I were to make an exit like his I would be rather happy than otherwise to have the chance to tell things to my distant descendants.

Why was Schnidi so careless of his belongings? They are strewn over about 100 m of the route just below the the pass. It is easy to see why the discards are preserved just here, in the former accumulation zone of a now-vanished glacier. But I cannot think of a reason why they should have been discarded just here.

Did Schnidi's religion oblige him, as thanksgiving for a successful crossing of the pass, to take off his trousers? More plausibly, perhaps our ancestors were about as careless as we are, losing stuff at random all along the route, but only the items buried by the ice have been preserved. On this interpretation, the Schnidejoch, when the valley was passable, was a moderately busy thoroughfare. In that case, why didn't the travellers cross by either of the passes lying a few kilometres to east and west, which are 200 to 500 m lower than Schnidejoch? Perhaps they did. Those passes may never have been in the accumulation zone of a glacier, in which case the Bernese Alps might have been even busier than implied by the Schnidejoch evidence.

Lastly, a question I have asked before, knowing that it won't be answered. What was Schnidi's word, or words, for the glacier over which he walked? He or she is concrete evidence for human interaction with glaciers in Roman times, and yet we have no record of a word for glacier in Latin.