YOU ARE HERE: LAT HomeCollections

Science File / An exploration of issues and trends
affecting science, medicine and the environment

On Shaky Ground

Despite High-Tech Tools, Predicting Volcanos' Behavior Is Still a Very Inexact Science


In the library of France's Mont Pelee Observatory on the Caribbean island of Martinique, there are several thick volumes detailing the known precursors of the May 8, 1902, eruption that killed 29,933 people in just two minutes.

At the time, there were no seismic networks to monitor the earthquakes that preceded the fatal event, nor were there geodetic measurements, tilt meters, strain meters or gas emission measuring devices.

Most of the observations before Pelee erupted involved changes in the fumaroles, or steam vents, on the mountain. One appeared as early as 1889, and there was quite a change in their number and volume in early 1902. The conclusions tended to be more subjective than the precise readings of quakes and other events taken today by an array of new instruments at threatening volcanoes.

But it is not clear that scientists in 1998 are much more certain about what precisely will happen with volcanoes than they were at Mont Pelee nearly a century ago. Be it the Soufriere Hills volcano on Montserrat in the Caribbean or the Long Valley Caldera at Mammoth Lakes, uncertainty continues to prevail.

Even when an eruption has already begun, it often remains impossible for scientists to decide how big it will be or how long a particular phase will last.

"It's very, very difficult, and we are uncertain," said Dan Miller, head of the U.S. international volcanic assistance team based in Vancouver, Wash. "Most of the time, even with all the instruments we have, all we can do is give local authorities trying to settle on the need for evacuations our best guesstimate."

On Montserrat, not far from Martinique, an ever-changing team of scientists monitors the ongoing eruption that began July 18, 1995 after 98 years of periodic earthquake swarms. They have instruments that can detect each rock fall, each earthquake and its particular type, and each flow of heated ash, rock and gas.

Developments are reported twice a day on the Internet at But perceptions of the future remain vague.

Recently, the scientific team said in a written report that the eruption might last for years and that it might be decades before the southern two-thirds of the island would be inhabitable again.

Similarly, scientists monitoring Popocatepetl, which has erupted sporadically since 1994, remain unsure about the future course of activity at that huge volcano near Mexico City.

Observations of erupting volcanoes are tough enough. But understanding volcanically restless areas, where an eruption may occur tomorrow or not for hundreds of years, is even more challenging.

This is what has scientists in a quandary near the Eastern Sierra town of Mammoth Lakes. At the 18-by-12-mile caldera, site of eruptions ranging back 760,000 years, one of the heaviest concentrations of scientific instruments of any place on Earth has been installed. Yet the scientists remain uncertain what is going on three to five miles under their feet.

They have stated that an eruption would probably be small to moderate, a conclusion based on deposits that show most eruptions there in the last 5,000 years have been that size.

Malcolm Johnston of the U.S. Geological Survey notes that in addition to its instrument network, the survey has intricate data in the form of graphs showing details of movement over time along various lines between monitoring stations. All measure ground deformation in the area of a vigorous earthquake swarm that has been occurring at Mammoth for seven months. The water levels in wells and temperatures in hot springs are also being monitored continuously.

Particular attention is paid to deformation. The uplift of terrain east of Mammoth Lakes has not reached the levels that marked eruptions in the Rabaul caldera in the South Pacific, or restless but non-erupting areas on land west of Naples, Italy. But David Hill, the Geological Survey's chief Mammoth monitor, has cautioned that volcanoes can erupt at varying levels of ground deformation.

Virtually daily, as the quake swarm continues, the Geological Survey has been updating its statement of conditions in the Long Valley caldera on the Internet at

The onset of the quake swarm, the scientists report, coincided with a slight increase in the rate of strain across the area involved, as revealed by geodimeter measurements, and a slight increase in a downward tilt toward the southeast detected by a long base tiltmeter near the Mammoth airport. In simple terms, this indicates that in the event of an eruption, debris could flow into Crowley Lake, a water storage point for the Los Angeles Department of Water and Power.

"These changes in deformation rate are consistent with a slight pressure increase within the magma body (molten rock) located at a depth of seven to 10 kilometers (four to six miles) beneath the area known as the resurgent dome," east of Mammoth Lakes, the scientists state.

Los Angeles Times Articles