When earthquakes started jolting Mount St. Helens in March, 1980, little infrastructure was available to monitor the mountain or warn of an eruption. As the quakes continued, scientists scavenged seismographs and surveying equipment to take the mountain’s pulse. Their agenda combined science (What is going on inside the volcano?) and public service (When and how should we warn residents and touristo-gawkers?)
After a couple of months of nerve-wracking foreplay, St. Helens exploded on May 18. Fifty-seven people died, mainly because the eruption started with a rare but phenomenally powerful sideways blast. Crusty old Harry Truman, who had become a media hero by refusing to leave his compound at Spirit Lake, died under a gigantic landslide.
St. Helens taught scientists a great deal about how volcanoes work — or at least, how that particular volcano works, says Steve Malone, who watched the events from a seismology lab at the University of Washington, where he now heads the Pacific Northwest Seismograph Network.
The experience, he says, was “like a baptism of fire” that “got us a huge leg up” on the difficult problem of predicting eruptions.
Volcanoes give four major categories of information that can serve as a basis for predictions:
Geologic history: What has this volcano done in the past? How regular and frequent were the eruptions, and what did they do?
Land deformation: Rising magma changes the surface in and near the crater. Broad deformation indicates that the magma is deep, while more focused deformation signifies shallow magma, and perhaps an eruption in the near future.
Gas emissions: Rising magma releases gases like carbon dioxide and sulfur dioxide. The timing and intensity of these gases hints at the magma’s location.
Earthquakes: Rising magma must bust through rock, and the resulting quakes may be the best single indication of magma movement.
Each of these factors became the subject of intense scrutiny as St. Helens rocked and rolled in spring, 1980. Just two years before, a U.S. Geological Survey report had predicted that the volcano might erupt before 2000. “It was a long-range forecast based on the geological record” of fairly regular eruptions, says Dan Dzurisin, a volcanologist with the Cascades Volcano Observatory.
Although there were plenty of warnings, there were no eruption forecasts as such in 1980, Dzurisin says. “Clearly we issued statements anticipating eruptive activity, but the exact timing was difficult to forecast. Even in hindsight, on May 18, 1980, there was nothing we were measuring that could have offered any short-term warning that it would start that morning.”
After the cataclysmic eruption, however, it was easy to find research funding on St. Helens, and the ensuing scientific activity paid off during a series of minor eruptions. “We began to see a pattern repeated … with increased seismicity and deformation of the crater floor,” Dzurisin says. “In most cases, a new lava lobe would extrude onto the surface. From December, 1980, to October, 1986, we successfully predicted each dome-building eruption, at least 18 of them, as far as three weeks in advance, although sometimes it was just hours in advance.” Domes are mounds of lava that form, usually inside the existing crater.
Using seismology only, scientists also predicted four of five explosive eruptions in the summer of 1980, Malone adds.
St. Helens made history, and not just for the giant landslide, says Dzurisin. “For the first time, were willing to step up to the plate and predict the style and timing of eruption events. It’s fair to say this was a watershed, and since then, it has become more common to make specific forecasts.”
St. Helens offered two types of lessons, says Dzurisin. “There are two aspects to an eruption forecast. One is pattern recognition. If you see the same thing happening repeatedly, you are in a position to forecast the next occurrence. To a large extent, that’s what we used for the dome-building forecasts at Mount St. Helens.”
A second aspect is understanding what’s going on inside an active volcano. “If we understand the processes, even if the pattern changes, we should be in a position to understand why they might have changed, and still provide some form of forecast,” Dzurisin adds. Although achieving such an understanding is “much more difficult,” it’s also potentially more helpful, since it might apply to other volcanoes.
Since 1980, the biggest advances in forecasting have come from seismology, says Dzurisin, who was at the Hawaii Volcano Observatory before St. Helens. “At HVO, we counted the number of quakes per day. …Now, instead of saying, we had 250 quakes yesterday and 324 today, we say” that groundwater or magma seems to be moving inside the volcano, based on the size and shape of traces on the seismograph.
A successful prediction
In 1991, the lessons of Mount St. Helens were tested in the Philippines, after the ground began shaking around Mount Pinatubo, a short mountain but productive volcano. A collaboration of U.S. and Filipino volcanologists watched, waited, and eventually made an accurate forecast. “There was no question that the learning curve took a huge jump at St. Helens, and they succeeded at Pinatubo, it was a dramatically successful forecast,” says Malone, who adds that he was not involved in Pinatubo.
The Philippine government evacuated tens of thousands of residents, saving massive casualties during the largest eruption of the 20th century. Although gigantic mud flows and ash deposits lead to the closing of two major US military bases, few lives were lost.
No evacuation warnings have been issued around Mount St. Helens, where, as we write, in October, 2004, magma has risen to a few hundred meters below the surface, and a lava dome is forming in the crater. “The fact that the deformation is very localized tells us that the immediate cause is very shallow,” Dzurisin said.
It’s unlikely that the surface lava flow is being driven by a monster flow deep under the volcano, Dzurisin adds. “If there was a large body of magma pressurizing it,” says Dzurisin, “it would definitely cause ground movement on the outer flank, but we are not seeing that.”
The mountain’s shape is also reassuring, courtesy of the 1980 eruption, Dzurisin adds. “There has been a change in the edifice itself. It was decapitated, so there is no potential for a large landslide, or the large, lateral, directed eruption. That’s off the table.