"Surveying on the summit of Merapi at dawn is marvelous," we quoted geophysicist Barry Voight as saying in the March 1990 Research/Penn State. "An array of dark cones -- Sumbing (9,000 feet high), Sundoro (7,500), Dieng (7,500), and Slamet (9,000) -- pierce the sea of clouds, marking the spine of Java. A golden band rims the horizon. From small villages far below rise faint cascades of silvery notes in richly harmonic chords -- gamelan orchestras greeting the dawn." At that time, Voight had just begun studying Merapi, seeking to test his mathematical method for predicting volcanic eruptions. In 1995, Voight and Penn State graduate student Kirby Young were sent to Merapi by the U.S. Agency for International Development and the U.S. Geological Survey to analyze a Noember 1994 eruption in which more than 60 villagers were killed by a "hot dust hurricane."

Called nuees ardentes (French for "glowing clouds") or pyroclastic flows (from the Latin for "fire" and "breaking apart"), these heavier-than-air blasts of hot dust, debris, and gas cascade down a volcano's side like an avalanche, uprooting and charring trees nearby, cutting off the tops of those further away, scorching what's farther out. "They can move as fast as 140 miles per hour," says Voight. "At Merapi, they take about three minutes to move the four miles to populated areas. Once they start, there is no chance of outrunning them. The heavy block and ash flow tends to follow the existing channels, while the associated ash cloud surge moves laterally beyond the channel margins."

Nearly a million people live on the slopes of Merapi, the most active of Java's 30 volcanoes. A quarter-million reside in the so-called "forbidden zone," which, Voight notes, is "a hazard zonation unfortunately without enforcement." The population density is one reason Voight chose in 1988 to study it rather than some other mountain on this South Seas end of the Ring of Fire. Instead of Pinatubo, for instance, some 1500 miles to the north in the Philippines. ("At that time," Voight notes, "Pinatubo was barely recognized as a dormant volcano. It wasn't even mentioned in the 1981 Smithsonian compilation of Volcanoes of the World.")

When Pinatubo erupted in June 1991, it lofted amazing amounts of sulfur and ash into the air, affecting the weather around the world for years. And although its slopes were less densely settled than Merapi's, hundreds of thousands were evacuated (including those at Clark Air Force Base and nearby Subic Bay Naval Station). "These things are not happy visitations on the poor people of these islands," Voight has said. "If there's any question about evacuating, it can be acted on immediately. The governments do not back away from their responsibility. Hazard management at Pinatubo was a huge success -- perhaps the highest point yet in volcanic disaster management worldwide -- but the problems are not all solved. Even with the evacuation, more than 600,000 persons lost their income, at least temporarily, and more than 100,000 homes were destroyed or damaged. Ultimately," he adds, "over a thousand died, many in collapsing structures destroyed by heavy ashfall, more by lahars -- flows of volcanic debris -- generated by heavy rainfall on the ash deposits, and still more by disease and exposure."

Java man was discovered on the east flank of Merapi. The largest Buddhist temple in the world is on the west flank. And at Merapi, Voight believes, careful monitoring of the mountaintop's shudders and shakes and swellings may enable scientists to mathematically predict when and where the rock structure will yield to pressure and the volcano will blow. Whenever he can, since 1988, he has been climbing to Merapi's summit, setting mirrors and meters in place, and waiting for the gold-tipped clouds to clear to take his measurements. (Once he caught graduate student Young reading a book called Experiences of Terror while they waited.)

But the November 1994 eruption he missed: His funding had run out. Some reviewers of his proposal to the National Science Foundation felt it wasn't "pure" research, since he was also training local scientists in his monitoring methods. Reviewers for development-oriented funding agencies, on the other hand, considered his project "too research-focused." ("And in any case," he adds, "the development agencies can seldom come up with funding before a crisis.") "All part of the frustrating side of this business," he says. "There is always this dilemma. Those of us who have personally experienced volcanic tragedies of the enormity of Armero, where 23,000 died in one hour, realize that our scientific mission must be aimed at solving real volcano problems. It is simply socially irresponsible to take advantage of the opportunities for pure research in developing countries without giving something back that is significantly useful to them."

The 1994 eruption, he knows now, "was caused mainly by gravity -- by the collapse of a portion of a thick lava wall that was pushed very slowly over the steep slopes" of Merapi. It was probably not an eruption that his mathematical method, which treats a volcano as an illustration of "rock mechanics" in action, could have predicted. "Most of the lava had already emerged, and the lava dome had been slowly growing for months," Voight explains. "Because there was no change in the rate of production of lava issuing from the vent, there was no change obvious in deformation patterns." That the eruption reached a populated area was also unforeseen. Says Voight, "A pile-up of debris against the wall of a channel in the lower slope enabled this fast, hot, bouldery dust cloud, the result of the collapse of the lava dome, to jump into a new river drainage area on the south flank of the mountain, an area that had been free of volcanic impact for a very long time. As a result of this combination of events, the hot ash hurricane reached a populated area where, as fate would have it, a wedding ceremony was occurring on the rim of a canyon.

"Everyone who was in the central area of the blast died," says Voight. "But in the peripheral regions some survived, and we are trying to identify what factors allowed them to survive." Of the 86 hospitalized, 23 pulled through.

According to Voight's research assistant, Ian Shelley, "Those who died sustained an average body surface area of severe burns of 74 percent, while for those who survived, the mean body surface area of severe burns was only 36 percent. The amount of surface area burned seems to be a good indicator the probability of life or death. Lung damage, caused by inhalation of hot gases and hot fine particles also contributed to mortality." Shelley and Voight found, however, that clothing -- even thin long-sleeved shirts and long trousers -- could decrease the burned area by 20 percent or more. "This decrease can make the difference between life and death," says Shelley. "Total burn area influences mortality from burn shock, and also influences delayed problems of vicious severity, such as multiple organ failure. If you can decrease your burn surface area by 20 percent, you have greatly improved your odds for survival."

The "lessons" to learn from this catastrophe, Voight and Shelley wrote in their abstract for the fall 1995 meeting of the American Geophysical Union, are that "individuals (including scientists) exposed to severe eruption risks on the volcano should wear protective clothing [and] carry surgical masks to reduce risk of trauma due to inhalation of hot particles."

"Just take off your hat and cover your face," says Shelley. "Those who had the presence of mind to do this at Merapi are still around to talk about the experience."

And you shouldn't run straight away (much as instinct urges you to), but to the side. Individuals (again, including scientists), "when facing clear danger, should run perpendicular to channels," they write. Only 30 feet, they found, separated "the zone of survival from the zone of probable death."

Remarking on the 1990 Research/Penn State article about his work at Mount St. Helens and Nevado del Ruiz (Columbia) volcanoes, Voight wrote, "It strikes me that, perhaps, my reasons for being 'in close' come off as too 'noble.' In fact, practically any volcano-bent individual between the Kattegatt and the Kuriles would have been delighted for the chance to work on Mount St. Helens' problems in 1980. Adventure was part of the reward, and along with the scientific opportunity, of course, came the responsibility to do things as best one could, in both science and hazard mitigation."

Of his first sight of Mount St. Helens, Voight remembers that "at dusk it sent up these marvelous orange columns." Of Merapi? It's "these marvelously aesthetic terraces of rice" on the hillsides down below. Even now, having seen once again their destructive power, he can concur with the old saw: Volcanoes, as they say, are places of terrible beauty.

Barry Voight, Ph.D., is professor of geophysics in the College of Earth and Mineral Sciences, 334A Deike Building, University Park, PA 16802; 814-865-9993. Kirby Young is a doctoral candidate in geosciences; Ian Shelley is a research assistant. Their work has been sponsored by the National Science Foundation, the U.S. Agency for International Development, and the U.S. Geological Survey.