|Your account||Today's news index||Weather||Traffic||Movies||Restaurants||Today's events|
Friday, October 29, 2004 - Page updated at 12:47 A.M.
Scientists edge closer to predicting volcanic eruptions
By Sandi Doughton
They said lava would probably reach the surface, and it did. They said not to expect anything like the catastrophic eruption of May 18, 1980, and so far, all of the outbursts have been modest.
But they weren't able to determine whether the mountain would unleash an ash-spewing explosion or a quiet outpouring of lava. They couldn't say when an eruption would occur. And now that molten rock has been oozing into the crater for nearly two weeks, they don't know how long it will continue or whether it will become more violent. "We're able to do a much better job of monitoring the changes that occur at volcanoes as they wake up, but we are still quite a long ways from being able to forecast the time, the magnitude and the character of an eruption," said C. Dan Miller, of the U.S. Geological Survey's Cascades Volcano Observatory.
Some experts say that could change within a few years.
Rapid advances in technology and innovative studies, including a bold project to drill into the heart of an active volcano, are poised to strip away some of the mystery that makes volcanoes so unpredictable.
In laboratories, researchers are re-creating the temperatures and pressures deep inside the Earth to study the gases that give molten rock its explosive pop.
Other scientists are wiring volcanoes with sophisticated new seismometers to better understand a type of rattle that frequently foreshadows eruption. And satellite radars promise to revolutionize the ability to detect tiny changes in a volcano's contours that signal the movement of magma.
"I'm pretty optimistic that all of this will translate into improved eruption forecasts quite quickly," said John Eichelberger, coordinating scientist at Alaska Volcano Observatory in Fairbanks.
With 500 million people around the world living near active volcanoes, more precise forecasts could save lives and reduce the social and economic toll of unnecessary evacuations, said Miller, who as leader of a USGS crisis-response team once urged 40,000 people in Ecuador to leave their homes then watched the volcano sputter but never erupt.
A magma's-eye view
Thanks to better methods for measuring earthquakes, ground swelling and gas emissions, volcanic forecasting has improved tremendously since 1980.
But the complexity of volcanoes and the inability to see their inner workings complicate the task.
Taking a page from Jules Verne's "Journey to the Center of the Earth," Eichelberger and his colleagues decided the best way to figure out what makes a volcano tick is to get inside one.
After nearly two years of drilling, they succeeded this summer, penetrating the magma conduit of Mount Unzen, Japan's deadliest volcano. In its last eruption, which ended in 1995, landslides and avalanches of hot rock killed 44 people, forced 15,000 out of their homes and destroyed $2 billion worth of property.
Diagrams usually depict volcanic plumbing as a round magma reservoir with a tube leading to the summit. In reality, the interior of a volcano is a network of cracks and crevices, said USGS volcanologist Willie Scott.
Understanding the details of that structure and the way magma moves within it will help in one of the most vexing challenges for geologists: Divining whether lava will come out with a bang or a whimper. The difference seems to lie mainly in the gas content of the magma, which depends partly on how fast it rises.
As molten rock moves toward the surface, dissolved water and gases begin to bubble out, like a carbonated soft drink. Magma that reaches the surface still loaded with gas and water vapor is much more likely to explode than magma that has lost its fizz. And the faster magma moves up, the fizzier it will be when it breaks through.
At Unzen, scientists will be able to calculate the rate of magma movement. They'll learn about the permeability of the rocks, the size of the cracks and conduits, and how they affect the magma's progress, Eichelberger said.
He and several other scientists around the world are also testing these ideas in the lab. "Twenty years ago, this line of research really didn't exist, " he said. "Now there's an enormous amount of work going on trying to understand the processes within magma rising through volcanic conduits."
Pops and thumps
One of the most powerful applications of the new insights will be to help explain what's responsible for the enigmatic thumps that seem to portend explosive eruptions.
Detected with seismographs, they're called long-period events because they can last up to two seconds. By contrast, most of the earthquakes when a volcano begins to stir are the short, sharp pops made by rock cracking like an ice cube in a cocktail.
USGS scientist Bernard Chouet first studied long-period events at Mount St. Helens and thinks they are caused by steam and magma rising through cracks and causing a resonant hum, like an organ pipe. He's also convinced they mean the system is pressurizing and getting ready to blow.
In 1990, Chouet predicted an eruption at Alaska's Redoubt volcano, allowing workers to be evacuated hours before mud flows wiped out their oil terminal. Several long-period events also rumbled beneath the Colombian volcano Galeras shortly before it erupted in 1994, killing seven volcanologists on a field trip.
The tragedy kicked off debate about the significance of the signals, but the past decade has clearly demonstrated their importance, said University of Washington seismologist Steve Malone.
"I don't know of any case in which a significant explosion took place, and there was good seismic monitoring, that you didn't see these events," he said.
But they're not a silver bullet, because they also occur at volcanoes that don't blow, Malone pointed out.
Scientists were alarmed earlier this month by a continuous rumble at Mount St. Helens. They warned that a large eruption could be imminent but all that followed were two small burps and the continuing lava flow.
Scientists still don't know enough about what triggers long-period events to be able to correlate them with the size of an impending eruption, Chouet said. Information from Unzen will help fill in those gaps, he predicted.
Satellites are also beginning to provide a view of volcanoes that can't be matched on the ground, said Dan Dzurisin of the Cascades Volcano Observatory.
Sensitive radar can peer through clouds and detect a slight bulging caused by the upward movement of magma, which alerted scientists to the swelling at the Three Sisters volcano complex in Central Oregon a couple of years ago.
With only two of the specialized satellites in space, the technique is limited. It wasn't helpful at Mount St. Helens this fall, because the 35-day orbit period was too slow to capture the rapid swelling in the crater before the lava flow started.
But with at least one new satellite set to launch soon, and more sophisticated instruments that can see through vegetation, it holds great promise, Dzurisin said.
Even at Mount St. Helens, researchers don't have all the instruments they want, and many break down in the harsh conditions. All of which underscores the role for a more low-tech approach, some scientists say.
The collective wisdom of veteran volcanologists is one of the most valuable tools in eruption forecasting, Malone said. As Mount St. Helens began rumbling in late last month, researchers spent hours on the phone, comparing notes from eruptions around the globe.
The USGS is beginning to assemble that know-how into a computerized data base anyone can use.
"Maybe someday we'll really understand what's going on inside volcanoes so we can just put out a few instruments" and know what they'll do, he said.
Until then, the expert data base can help geologists around the world make their best guesses.
Sandi Doughton: 206-464-2491 or email@example.com
Copyright © 2004 The Seattle Times Company
Home delivery | Contact us | Search archive | Site map | Low-graphic
NWclassifieds | NWsource | Advertising info | The Seattle Times Company
Back to top