February 9, 1993

When a Volcano Turns Deadly for Those Studying Its Moods

By WILLIAM J. BROAD

is legs broken and jaw wired shut, a battered survivor of one of the worst disasters in the history of scientific fieldwork has recounted from a hospital bed how six colleagues died while searching for better ways to predict when volcanoes are going to blow up.

The six were killed on Jan. 14 as they gathered chemical and gravitational clues in the cone of the Galeras volcano, stretching 13,680 feet into the cold, thin air of the Colombian Andes. The ground suddenly began to heave. With a deafening roar, the volcano exploded in a riot of incandescent boulders and lava and ash, some thrown miles high.

The blast crushed or burned to death six scientists in or near the crater. The lone survivor among the group was Dr. Stanley N. Williams, 40, a volcanologist at Arizona State University who is married and the father of two small children, aged 7 and 4.

Dr. Williams, standing on the crater's eastern rim, watched helplessly as two colleagues inside the crater and three on the opposite rim were caught in the upheaval.

"I said to myself, 'I don't want to die. I don't want to lose my wife and kids,' " he recalled in a telephone interview last week from a hospital in Phoenix. "I turned and ran as fast as I could. I didn't make it very far, only about 20 meters below the rim."

His skull fractured and legs broken, Dr. Williams fought to retain consciousness while crumpled on the slope so he could watch for falling boulders, dodging them as they dropped.

The catastrophe struck as 90 scientists from 15 countries met at a conference in Pasto, Colombia, to study the Galeras volcano as part of a United Nations effort to reduce the death toll from natural disasters. The volcano had been judged likely to have a large eruption this decade and to pose a significant threat to humans.

Pasto, with a population of 300,000, lies at the foot of the volcano's eastern slope a few miles from the crater, and a total of 400,000 people live on the volcano's flanks. Galeras is today the most active of the Colombian volcanoes, having had major eruptions with lava flows in 1936 and 1945 and a number of minor ones since reawakening five years ago, including an eruption last July.

Dr. Williams led the American contingent at Pasto, which is about 375 miles southwest of Bogota in the Andes near Ecuador. The disaster struck during a field trip by the 90 scientists to examine the volcano and collect data.

Although the absence of discernible warning underscored the primitive state of eruption forecasting, the scientists now recovering from the disaster say measurements made at the Galeras crater hold significant promise for the future. Chemical clues in vented gas and subtle shifts in the volcano's overall gravitational field, they say, may eventually give scientists a new kind of alarm system.

"They could tell us when we're getting into a dangerous period and more surveillance is needed," said Dr. John Stix, a volcanologist at the University of Montreal who helped organize the meeting.

In theory such warnings could save many lives. The eruption in 1985 of Nevado del Ruiz in Colombia killed some 25,000 people.

Red Glow in Night

Volcanoes are fickle. They have distinct personalities and all too often show little similarity in their patterns of discontent. While Galeras exploded last month with no warning, Kilauea on Hawaii sent all sorts of clear signals before erupting after midnight on Jan. 2, 1983, beginning one of the longest volcanic eruptions in recorded history.

The ground around Kilauea shook and became distended as molten rock forced its way toward the surface. Finally, after nearly 24 hours of warmup exercises, earthquakes were replaced by rhythmic vibrations, a sign of gushing lava. Sure enough, the nighttime darkness around the volcano was lighted by a red glow.

In 1989, one year after Galeras reawakened, the Colombian Government established a geologic observatory at Pasto to monitor earth tremors, and ringed the volcano with a network of seismic sensors. It also set up stations to measure rises in ground height, which are thought to occur as magma (the molten rock that turns into lava) accumulates in reservoirs beneath a volcanic summit. Such monitoring, however, gave no warning of last month's fiery explosion.

Among the aims of the weeklong Galeras meeting, which began on Jan. 11, were to discuss bolstering such traditional methods with newer and more sensitive ones.

A day after the meeting opened, Dr. Geoff C. Brown, a geologist at the Open University in Britain, presented a paper on how infinitesimal changes in gravity around volcanoes could help reveal the upward movement of magma. When a volcano becomes distended, it moves ever so slightly, often just inches, away from the earth's center of gravity. This produces a very small reduction in the strength of the gravitational field at the volcano that can be measured by instruments of great sensitivity.

Another stratagem discussed at the meeting was tracking changes in the composition of vented gas. As magma rises near the surface and undergoes a huge drop in pressure, dissolved gases come out of solution (like carbon dioxide bubbling out of a bottle of soda or seltzer). At first the lightest gases predominate -- carbon dioxide, followed by water vapor.

Appearance of Heavier Gases

But if magma sits in a reservoir, getting relatively old, the rush of light gases is eventually replaced by heavier ones like sulfur dioxide and hydrogen chloride.

Thus, geochemists believe that monitoring the composition of vented gas might announce the arrival of fresh magma -- and warn of increased danger of an eruption.

Both chemical and gravitational clues were being gathered during the morning field trip to Galeras on Jan. 14. Everything went smoothly at first amid the cool fog swirling about the volcanic peak. More than a dozen scientists entered the volcano's crater to collect data, as well as some journalists and tourists.

Evacuated bottles were used to sample gases. Dr. Brown of the Open University took gravity measurements on a portable meter.

By noon or a little afterward, half the scientists had finished their work and left the crater. Four Colombians stayed -- Dr. Jose Arles Zapata, a geochemist at the Pasto observatory; Dr. Fernando Cuenca, a geophysicist from Bogota who had conducted a magnetic survey of the volcano; Dr. Nestor Garcia, an industrial chemist at the National University, and Dr. Carlos Trujillo, a community college teacher in Pasto who liked to use the volcano as a classroom.

Also remaining were Dr. Brown of the Open University and Dr. Igor Menyailov, 67, a Russian scientist from the Institute of Volcanology in Petropavlovsk on the Kamchatka peninsula.

Onset of Disaster

The Russian relaxed at the bottom of the crater, happy with the day's work, smoking a cigarette. "He had never been to South America before," Dr. Williams recalled. "He was excited."

Dr. Williams stood on the crater's eastern rim as the remaining scientists left or got ready to leave. "I was talking to them about what they were doing and whether they were pleased with the data," he said. "Igor stood there smoking. He was going to climb up to me."

Suddenly, the crater's floor began to lurch and the top of the volcano exploded in a frenzy of hot rocks and lava, killing six of the scientists as well as three nearby tourists.

"I heard this huge boom, and then rocks the size of televisions started falling around us," recalled Dr. Andrew McFarlane, a geologist at Florida International University who had got beyond the crater. Dr. McFarlane suffered a broken foot, bruises on his legs and badly burned hands from climbing over burning rocks.

Dr. Williams, fleeing the crater's rim, pounded by flying rocks, ran as far as he could down the volcanic slope before his broken legs gave way. He took shelter from the weakening eruption behind large rocks. After an hour, a second volcanic blast hurled aloft new boulders that he successfully dodged.

Through it all the earth shook and rumbled, leaving geologists in the area fearful that the volcano had entered a new, more dangerous phase before a major eruption.

After some two hours on the crater's slope, Dr. Williams heard in the distance his graduate student, Marta Lucia Calvache, who is the acting director of the Pasto geologic observatory, and another woman calling out to see if anyone was alive.

'They Saved My Life'

The two women, atop a ridge overlooking the volcanic crater, came down with a stretcher and hauled Dr. Williams to safety. "They saved my life," he said, choking back emotion.

He was flown by helicopter from the volcano's side to a hospital in Pasto, and later transferred to one in Phoenix. He expects to be out of the hospital later this week.

The eruption of Galeras last month turned out to be the largest there in five years, as measured by the height of its volcanic cloud and the release of seismic energy. But it did not pump lava down the side of the mountain, which could have resulted in a major catastrophe.

One of the tale's many twists is that several scientists from the United States Geological Survey who had planned to attend the meeting and would have probably been in the crater during its eruption were barred at the last minute by the State Department, which feared political violence linked to Colombia's drug lords.

A memorial of sorts to Dr. Brown of the Open University appeared in the current issue of Nature, the British science journal. A paper of his, written with four other scientists, presented gravity data from the monitoring of Mount Etna in Sicily and concluded that measurements of ground waves and deformation had missed critical motions of magma near the surface.

"Only gravity changes," Dr. Brown and his colleagues wrote, "can detect the mass increase."

Why did the new techniques tested at Galeras fail to warn the volcanologists of danger? Dr. Williams said there might well have been chemical clues, but these will only be apparent after the samples are processed in the United States.

The goal of the gas program, Dr. Williams added, is a portable unit that could be left near a volcanic vent to analyze gases automatically every few hours and radio the results to distant scientists. The feasibility of such a device is being investigated by Transducer Research Inc., a small company in Naperville, Ill.

Lack of Warning

Another possible reason for the lack of warning, said Dr. Stix of the University of Montreal, is that the blasts involved no major movement of magma but simply the release of pent-up gases. If so, he said, "the volcano was less dangerous after the eruption than we had initially thought."

Dr. Williams, though shaken by the death of his colleagues, said he was ready to go back to Galeras despite the obvious dangers. He noted that volcanology was an inherently high-risk profession in which on average of one scientist dies on the job every year or so.

The Galeras disaster, he added, might have a silver lining if it focused global attention on the problems of volcanic forecasting and resulted in new funds being devoted to avoiding disasters.

"The important thing," he said, "is that people follow through with this research and make an impact."

Return to the Books Home Page