by Lev DESINOV, Cand. Sc. (Geogr.), RAS Geography Institute, Sergei DESINOY, research scientist, same institute

On 20 September 2002 the Kolka glacier suddenly avalanched 50 km from the capital of Northern Osetiya. More than 150 mln m ³ of mud and stones rushed down wiping off everything on their way, killing almost 130 people, destroying 15 houses of the settlement of Nizhni Karmadon, two recreation camps, a sanatorium and blocking spa wells and motor roads in the hollow. The total damage was estimated at about 18 mln dollars.

What is the cause of the calamity which has brought about such tragic effects? Before we suggest our version of an answer, let us point out: what happened is a unique event in the history of glaciology. No glacier in the world has ever left its bed completely.

The Kolka glacier lies along the steep northern slope of the Kazbek-Djimaray plexus of mountains at the head of the Genaldon river. The dominating position here is occupied by the Maili glacier, one of the largest in Northern Osetiya (6.8 km ² ). It is mainly fed by granular-snow fields of Kazbek and is therefore brightly white in appearance being a good tie for space photography. Left of Maili, in a deep car, the Kolka glacier is located (2.5 km ² ) concealed by a sheath of stone moraine ^* , therefore it is hard to locate from above. Remarkably, it was explored better than other glaciers. Expeditions would not leave it for years. Dr. Rototayev, a glaciologist of the RAS Institute of Geography, worked there for 8 years. In early 1980s in his book "Study of the Kolka Surging Glacier" he predicted the future calamity.

Anyway, the Genaldon accident was anything but surprise. It was preceded by a general collapse of the system of monitoring hazardous mountain processes in Russia. Some disturbing facts were simply ignored. Thus, in the course of the Hurricane program on the Russian segment of the International Space Station (ISS), the Kolka glacier with signs of approaching shift was captured by the KODAK DCS 760 digital camera and photographed by cosmonaut Korzun 38 days before the disaster. But the online satellite data were never transmitted back to earth. A reason for that has been that the US Mission Control Center providing integrated mission

^* Moraine-deposits accumulated by glaciers in the course of drift and bed exaration. - Ed.

Articles in this rubric reflect the opinion of the authors. -Ed.

Pages. 66

control restricted the use of ISS radio channel for the Hurricane.

The tragedy has given rise to swarming concoctions presented as all kinds of "conclusions" and "versions", each new one more incredible than the former: "the avalanche was triggered off by fireworks staged by moviemakers"; "the catastrophe was caused by the events in Chechnya"; "main glaciers collapsed hitting the Kolka off its bed" and so forth.

The latter version is most popular. If we take it for granted, then we shall have to agree with at least three conclusions. First, the descent of a large mass of brittle overfrozen laminar ice down a slope less than 40 in tilt is estimatedly able to hit off its bed a glacial body with the mass of about 200 million cubic meters and up to 150 meters thick. However, if you make one or two mounts in high lands you will see that in reality at such collapses just small ice pieces reach the glacier surface, and the entire tumbled mass covers it as ice pellets.

Second, the population of Caucasian valleys should be urgently evacuated and similar measures should be taken by governments of many alpine countries. The fact is that in populated alpine localities throughout the world key glaciers are overhanged by their smaller cliff counterparts.

And, finally, third, dismaying, conclusion: if this approach is to be taken as the only right, from now on teams monitoring hazardous alpine glaciers should be trained to look up at the mountain ridges, rather than at glaciers themselves.

So, what did really happen at the Kolka? Our analysis is based on the 40-year experience of studying dozens of similar objects amassed by glaciologists of the RAS Institute of Geography since the time the shift of the Medvezhiy glacier in Pamir was first studied. We largely drew on the results obtained through interpreting satellite images of Pamir, Tien Shan, Karakoram, Southern Andes and Alaska mountain systems.

Any glacier on earth is a natural body of frozen water of atmospheric origin which possesses the property of auto-motion. The latter presents its key characteristic feature. A glacier or its part which has lost it is termed dead ice. Therefore, such natural formations do not include icebergs, ice buildups in deep niches or ice lenses in permafrost, etc. Glaciers only appear if perennial supply of solid precipitation exceeds melting of snow and ice.

Such conditions exist all over the globe. All points above its surface where this rule is observed in their totality form what is termed chionosphere. Even in tropics it does not reach above 8 km going down to the ocean level further north. In mountainous regions chionosphere is manifested by snow line. Above it, in the accumulation area, glaciers build up, and below, in the ablation zone, disburse the mass.

At the Kolka the snow line lies below the Kazbek plateau, between the main glacier and its cliff-hanging counterparts which feed the former with slumps of their bodies. Other contributing factors of the Kolka are solid precipitation, snow-glides, snowdrift and collapse of rock. Being a natural debris transporter, any glacier accumulates it and takes down.

Experts term it thawing out of debris material and formation of surface moraine. What is peculiar of the Kolka is just a large percentage content of rocks in the input of matter. Its body is literally littered with them.

Usually glaciers have an extensive network of water streams in their bed. Most often, escaping it, they conflow into a single stream coming out of a grotto-an arch-vaulted tunnel. But water may find another more natural way out, then a new arch-vaulted tunnel appears in the glacier, and the old one is drained.

None yet observed a waterway network in a glacier bed. Even radio probing is unable to produce a complete picture. What is unique about the Kolka is that its upper area is a flatly sloped surface with a rocky outcrop at least 15m high. That means that in the autumn of 2002 the sub-ice hydro-network on the top of the glacier most likely presented not streams but a whole lake.

Presently, most of world glaciers are in the phase of recess. Retreating deeper back to their valleys, they leave distinct traces, including fosses or scars on slopes of valleys, but most impressive are moraine deposits. The most conspicuous of those in the Genaldon valley have allowed to reconstruct the former shifts of the Kolka in 1902 and 1969. Having reviewed the materials, Dr. Rototayev has referred the glacier to a surging type with the cycle of 65 - 70 years. The main distinctive feature of such glaciers is instability of dynamic conditions determined by the excess of ice supply over its debit and the morphological structure of the natural formation and the adjacent valley.

In surging glaciers smooth melting of ice is periodically, at roughly equal intervals (varying by object), succeeded by quick one. That happens irrespective of any environmental factors (e.g., abrupt climatic changes) and is triggered by instability of dynamic processes in glaciers themselves. As for earthquakes, they seem to be capable of causing faster accumulation of the glacier critical mass, but can hardly trigger off its shift. This fact is commonly hard to perceive for laymen.

The increased interest to surging glaciers is stirred by the threat they pause to population and economic facilities. The fast advance of an ice mountain often produces supply lakes in a blocked outer valley. Breaking through the block they cause disastrous floods. On other rarer occasions a similar shift of an ice mass is accompanied by separation of its part which rushes tumbling down.

As mentioned, Dr. Rototayev estimated the Kolka surging cycle at 65 - 70 years. That is, its next avalanche should have been expected in 2030 - 2035. But the critical mass accumulated at the Kolka by 2002. Why?

The most straightforward explanation, in our view, is that the glacier's surging cycle proved to be twice as short as expected. We do not have sufficient data about the glacier condition back in 1935 - 1940 when another microshift might have happened. Further, sometimes surging fades out inside the ice mountain contour instead of going out. Or the length of a glacier may increase so insignificantly that such an event passes unnoticed. E.g., the world's best

Pages. 69

studied surging glacier Medvezhiy which blocked the valley of the Abdu-kagor river at Pamir in 1963, 1973 and 1989 with rather an impressive ice dam, in its descent of 2001 all of a sudden stopped just 50 m short of its buttress causing no hindrance to the water flow.

We do not question the accuracy of Kolka surging cycle as calculated by Rototayev. It is just that a number of factors expedited its avalanching. One of those may be still obscure processes of thermal behavior within the nearby Kazbek volcano.

There is still another possible, purely geological factor: the positioning of the ice and rock collapse zone strictly along the route of a powerful quasi-latitudinal fracture crossing the Kazbek-Djimaray system.

Faster glacier avalanching, i.e., abrupt surge of exogenous factors on a locally limited area of the plateau cliff might have been indirectly stimulated by a series of earthquakes that took place in the Caucasus in August 2002, or reverberations of remoter disturbances of the earth crust. In that period, i.e., from 28 August to 5 September alpinists of Krasnodar registered continuous ice and rock falls. In other words, accelerated buildup of the critical mass started at least 4 weeks before the disaster.

We propound: the cause that triggered the mechanism of Kolka slump were the earthquakes that occurred on July 14, 2002 at 00:53 a.m. and August 22, 2002 at 12:25 p.m. in 70 - 75 km north-west of it. We believe it is these subsoil shocks that provoked the gradual fall of masses off cliff glaciers.

The area of collapsed ice of about 200,000 m ² and the glacier thickness of 15 - 25 m, in a tentative estimate, correspond to the mass of 3 - 5 mm tons. It is much harder to calculate the volume and weight of slumped rock. One way or another, we agree with the opinion that about 8 million tons of matter fell on the Kolka from the southern wall. But it happened not all at once but was stretched in time throughout almost a month. And the slope produced not a single grand collapse of a cliff glacier but a lengthy fall of mass and rock.

In Rototayev's estimate, in normal conditions the glacier had an annual positive balance of 1 - 1.3 mln tons. Hence, within one month it accumulated a surplus of matter equal to the 5 - 8 year volume. In the state of commencing shift, when the outflow area pressure caused the activation of the glacier mass, Kolka went into the most dramatic phase of surging.

The start of activation should be dated back to the second half of the summer, since the image obtained by cosmonaut Korzun shows that already by August 13, 2002 the ice tongue featured four consecutive arc waves. Having gained momentum for such an extraordinary reason, the Kolka shift further followed the standard pattern until the bed water came into play.

Usually, in the first phase of surging ice driven down by gravity comes in several rounds. The active upper glacier zone develops an expressed bulged cross section, while the passive zone remains concave, masked by the surrounding terrain. Incidentally, the latter of the two mentioned deciphering signs is also explicitly present on Korzun's satellite image. Had there been professional observers on the glacier in the second decade of August, they would have been sure to notice these peculiarities.

As waves of activity hit the bottom point of the glacier, the latter separates from the bed and starts to shift. Since ice belongs to the category of the so-called viscoplastic materials, it cracks longitudinally and cross-wise. New supplies of ice mass boost the splintering, the glacier outer layer splits into thousands of prismatic blocks, and a thick network of cracks covers its face. This third, most conspicuous indicator of surging was never taken notice of, although photos of Krasnodar alpinists taken in early September already show first such cracks in the rear zone, and large ones at that.

Softened like that, the glacier begins to rise like dough, trying to squeeze its mass into bank moraines. This is the manifest indicator that the surging glacier is activated. Then its mass gets over bank moraines filling flank shallows along the slope. Retreating, the ice jumbo will deposit a part of its mass there. Such ice patches really trail along the Kolka left slope for hundreds of meters. They will melt completely with time.

The next significant indicator of a surging glacier revitalization is the temporarily drying out bed of the river outflowing from beneath it with the subsequent discharge of water and ice pulp into it. It was also registered by alpinists. On August 28 the torrent burst out from the river head and stopped at the Upper Karmadon spas.

And now it is time to recall the water flooding the Kolka bed. It was abundantly flowing from the southern slope where the ice was melting faster. And there is another significant factor which should not be overlooked: the past three years in the Genaldon valley were extremely rainy. The bench in the bed which was blocking drain into the valley served as a safeguard. Starting shift, the glacier cut a broad trench in the dike, thus getting abundant "lubrication". This water cushion largely facilitated the Kolka slump.

We shall never know just how many days, hours or minutes before the disaster the advancing ice reached the sharp bend of the valley and came to a halt. Generally, it is extremely dangerous when a surging glacier encounters obstacles on its way: running into them it pumps up muscles.

Another handicap encountered by the Kolka was the need to squeeze into the valley through a very narrow cramped scissure. It was extremely hard to overcome, but the glacier successfully did it. Gaining power at the bend and channeling it into this lock, the glacier took its start. Escaping from the gaping in the rock, the viscous mass of water, ice and stones swept through the Genaldon valley at a monstrous velocity, crashing and smashing everything in its way. A glacial disaster of such magnitude has never been observed in the mountains of this country. That makes us stop and think about the nature of the calamity and draw the necessary conclusions for the future.

Permanent link to this publication: