On April 21, 2000, Jeff Miller was skiing near Field, British Columbia, when the unthinkable happened. "The avalanche sucked me down like a whirlpool, " he writes. "The last thing I remember was flying through the air with my hands above my head and a sea of white all around me. The snow had turned to cement and even though I was only buried about a foot deep I fought to free myself." Although badly injured, Miller was lucky enough to survive to tell his terrifying story.

The rescue statistics for avalanched people are grim. The chances of survival under the snow fall dramatically after just 15 minutes, according to the Colorado Avalanche Information Center; only one avalanched person in three is found alive after one hour. Training and wearing beacons increases the probability of survival, but, experts say, preventing accidents is still the key to saving lives.

Every year, snow slides kill about 20 people in the U.S. and about 200 worldwide, mostly skiers and snowmobilers. Fatalities are only part of the problem, however. Avalanches cause billions of dollars of damage annually to buildings, roads and livestock and threaten many mountain communities. In Switzerland, for example, an estimated 65 percent of the population live in areas that are at risk for avalanches.

Most countries have set up networks of centers to monitor and forecast snowslides, which has helped to limit the number of victims during the past 20 years. When the centers warn of a high risk of avalanche, skiers are advised to steer clear of dangerous areas, and authorities can take the necessary steps to protect people and infrastructure. Passive defenses against avalanches include evacuating villages, bridges, roads and ski resorts if they are threatened by an avalanche. Many infrastructures feature concrete barriers that halt or divert possible slides. Workers also block the formation of big slides by setting up fences and nets that break down the snow mass. They even provoke small avalanches with cannons or explosives in evacuated areas, preventing bigger slides from occurring spontaneously when people are in the vicinity.

There are two types of avalanches: sluffs and slabs. In sluffs the snow tumbles loosely down a slope, with the mass growing as it gathers more snow on the way. Slabs are the most common and deadly form of avalanche, in which entire layers of the snowpack break loose and slide, burying everything in their path.

Several factors affect the likelihood of an avalanche, including temperature, weather, the force and direction of the wind, and the steepness and orientation of the slopes. The single most important factor, however, is the condition of the snowpack and the way it has developed over the season. Although we usually see only the surface of the snow, the cause of an avalanche often lies several feet below, in the bottom layers of the snowpack.

"Like a cake, the snowpack is made up of many layers with different firmness and mechanical properties," explains Mauro Valt, a researcher at the avalanche center in Arabba, Italy. The conditions of the snow are ever changing, depending largely on its building blocks: the flakes, or snow crystals. As the Inuit well know, flakes can take on endless forms. The World Meteorological Organization has catalogued the 10 types that are most relevant to snow monitoring. Ideally, when the temperature is low and the wind is weak, flakes deposit on the ground as the classic star-shaped crystals, resulting in the dry, soft layer that skiers favor. This situation does not last long, however. Subsequent snowfalls then add new layers, which may vary depending on the weather conditions. Changes in temperature, sun, rain and the compression of the snow mass may transform the star-shaped flakes into a variety of grains with different shapes and properties.

When the snow refreezes after a warm day, grains become round with a diameter of about 0.5 millimeter, forming very stable layers. When the temperature drops dramatically for many days, on the other hand, grains in the lower layers grow and become pyramid-shaped. The resulting snow, called depth hoar, is quite unstable, providing a weak base for the upper layers. In this situation, even minor pressure--the weight of an animal, or skier, for example--can trigger a devastating slab. Indeed, in nearly all reported cases skiers or snowmobilers have been victims of avalanches accidentally provoked by themselves.

For avalanche forecasters, up-to-date data about the snowpack are as vital as the temperature and the weather, but harder to obtain. Whereas high-tech weather stations can be run automatically, snowpack monitoring requires considerable low-tech and difficult fieldwork. Every day, experts from avalanche centers monitor the snowpack surface in different areas and dig tens of snow pits, the only way to view cross sections of the snowpack and its various layers. Workers check the size and shape of grains in each snow pit and test the stability of the snow. Bad weather and risk of avalanche often hamper fieldwork. Scientists are thus looking for new ways to obtain real-time information about the snow that will make avalanche forecasting safer, easier and more precise.

A fresh look from above
A May morning of 12C is pretty warm if you are at the Ny lesund arctic research base in the Svalbard Islands, a one-hour flight from the North Pole. In the flat, icy desert surrounding the research base, Mauro Valt is digging a snowpit. In a matter of minutes, hes down in the pit to the shoulders, scraping some snow from the layers and scrutinizing the flakes with a magnifier. Meanwhile a group of scientists from the Italian National Research Council (CNR) unwrap light sensors and laptops. There is no risk of avalanche in this locale. Instead what brought the Italian team to the arctic is the goal of enlisting satellite technologies in the fight against avalanches. "The arctic environment is an ideal open lab for our tests because it has huge snowy flats, no pollution and during summer the sunlight is round-the-clock," says Rosamaria Salvatori, who leads the Italian team.

Before leaving the base camp, the researchers downloaded fresh satellite images of the area that reveal precisely how much sunlight the snow currently reflects. "The amount of reflected light depends on the snows thickness, the content of water and on the type of crystals that it contains," Salvatori explains. Over the course of years of fieldwork, investigators have been correlating the satellite data with the type of snow found on the ground. They have discovered, for example, that the patterns of reflection in the infrared spectrum differ among fresh, wet or refrozen snow. Crystals that are star-shaped or that have been remodeled by the wind also reflect light differently. "To a certain extent, we can now guess from the satellite data which type of snow is on the ground," Salvatori reports. Even if satellites can see only the upper four to five centimeters of the snowpack, they can tell whether lots of fresh snow has just fallen over an area, or indicate the presence of superficial hoar--two conditions that increase avalanche danger. The Italian team hopes that one day avalanche centers could use satellites to monitor remote areas in real time, limiting fieldwork. Another team at the University of Colorado in Boulder is working independently on snow satellite sensing, with similar results.

More research is needed before these systems can be applied to avalanche forecast, however. "It will need a lot of tuning up before we can transfer the results we obtained in the arctic to the mountains, where we have to take into account slopes, pollution, rough terrain, vegetation and shadows," Salvatori cautions. Nevertheless, researchers are optimistic. "Satellites will not stop us completely from digging snow pits," Valt remarks, "but we are confident that they may ease our daily routine and risk and provide a more precise forecast of avalanches."

Sergio Pistoi (www.sergiopistoi.com) is a freelance science and medical journalist based in Rome, Italy. In May 2003 he traveled to the Ny lesund research base at the Svalbards with a grant from the Italian Association of Science Writers UGIS.