For those unfortunate enough to be buried underneath a layer of snow, lack of air kills long before the cold takes serious effect: either the victim quickly exhausts all the air available in the vicinity, or external pressure inhibits the mechanical ability to breathe. Statistics indicate a 90 per cent chance of survival if victims are found within 15 minutes of being covered. Over the next 20 minutes the chances of survival fall to 30 per cent, and only one in 10 of those buried longer than an hour survives.
Time is therefore of the essence for survival. The traditional method of locating victims of an avalanche is by the use of prodding poles pushed down in likely locations until something hard is felt. More recently, electronic beacons are sometimes carried by skiers, so that a radio signal will guide searchers in the right direction. But even nowadays it is found that the most efficient aids are sniffer-dogs, who proceed unerringly and with little hesitation to their human destination.
Ideally, of course, we ought to be able to avoid such situations altogether, by predicting when an avalanche will occur on a particular slope. This, however, has proved to be an elusive goal, although a number of methods are currently the subject of research.
One, for example, is an "analogue" technique similar to that used some years ago for longrange weather forecasting. At a vulnerable location, avalanches of the past are correlated with the weather conditions of the preceding days or weeks, and a database built up. Also available to the computer is the physical profile of the mountain-slope, and details of the current depth, and hence the weight, of snow. Every day during the season when an avalanche might possibly occur, the recent history of the local weather is fed into the computer, and if it finds a match with any previous occurrences, it may well be that there is a serious threat.
Another methodology is more direct. A hole is dug on the vulnerable slope, and an expert assesses the risk of avalanche by observing the texture of the layers of snow in the first six feet or so below the surface. And a third technique carries out the same assessment more scientifically: ice crystals of different kinds absorb or reflect "white" light in different ways, so if a sample "core" of snow is extracted from the slope by means of a hollow tube, the characteristics of the different layers can be deduced by shining a light upon it and using an instrument called a spectroscope. This knowledge, in turn, can be used to predict the likelihood of avalanche.