Volcano: In the Belly of the Beast: Scientists know what makes a volcano blow but still cannot say when

(2 of 3)

Employing a sort of tectonic checklist, geologists divide volcanoes into a few large families, ranging from the potentially explosive subduction type, like Nevado del Ruiz, to the milder plume variety, like the famed Kilauea volcano in Hawaii. In the subduction volcano, the magma is composed largely of andesite, which is a melt relatively rich in silica, water vapor and other gases. These gases are trapped until the magma approaches the surface, at which point, explains Richard Hoblitt, a USGS volcanologist in Denver, the pressure is released and "the gases tend to escape explosively." Another renowned example of an explosive volcano caused by subduction forces is none other than Mount St. Helens. Alfred Anderson, a geologist at the University of Chicago, says the two are almost identical.

Kilauea, on the other hand, frequently belches forth bright orange ribbons of glowing lava (the word for magma once it reaches the surface) in eruptions that are relatively gradual and benign and seldom cause loss of life. This plume magma contains less silica, is lighter than subduction magma, and allows expanding gases in the incandescent mixture to escape continuously as it moves toward the surface.

The Colombian eruption was made devastating by the large amounts of ice on the volcano's cap. "Nevado" is a Spanish word for snow, and though the mountain is near the equator, it is so tall that it is topped by snow and partly sheathed by glaciers year round. When the volcano blew, it tossed out hot material of varying sizes, from tiny grains to large rocks. The glowing debris hit the ice and melted it, causing torrents of scalding water to cascade down the slopes, mixing with volcanic debris and forming lahars, or mud-flows. That slurry, says Kevin Rodolfo of the University of Illinois at Chicago, "goes tearing down the banks of the volcano" at speeds of 30 m.p.h. In prehistoric times lahars raced down the sides of Washington State's Mount Rainier and traveled as far as 50 miles. "That's where the problem comes in," Rodolfo notes. "People don't expect to be affected 30 miles from the summit, and there's no time to warn them."

Being able to warn people to get out of the way long before the lahars descend is one goal of volcanologists. Right now the business of volcano predicting is a hodgepodge of high technology, empirical evidence, common sense and a certain amount of luck. As with earthquakes, geologists can often say that something will occur; the question always is exactly what and when. For long-range predictions, scientists try to reconstruct the history of a volcano, using such techniques as radiometric dating of volcanic deposits and studying tree rings (trees damaged by lahars will record the event with thinner, abraded rings). Says Hoblitt: "A volcano's behavior in the recent geological past is our best guide to behavior in the near future."