Listen to a Volcano Scream Just Before It Erupts

Scientists are used to a seismic brouhaha rising up from the molten depths before a volcano blows its top. But the moments before the 2009 eruption of Redoubt Volcano were different. The Earth began to scream. Using sonified seismic data collected in the days surrounding the eruption of Alaska’s Redoubt Volcano in March 2009, Stanford scientists now offer an intriguing explanation for the seismic weirdness that preceded the violent event.
Redoubt Volcano caldera
Redoubt Volcano on March 31, 2009. View to the east of the summit crater of the volcano, heavily covered with deposits from recent eruptions, many of which were preceded by harmonic tremor.(Game McGimsey)

Scientists are used to a seismic brouhaha before a volcano blows its top. But the moments before the 2009 eruption of Alaska’s Redoubt Volcano were different. The Earth began to scream.

Using sonified earthquake data collected in the days surrounding the eruption of Redoubt Volcano in March 2009, Stanford University scientists now offer an intriguing explanation for the seismic weirdness that preceded the violent event. The research, led by Eric Dunham and published this week in Nature Geosciences, describes how special harmonic microquakes may result from small, highly energized faults scraping together like fingernails on a chalkboard. Dunham’s new work builds on several years of volcanic research and offers the best explanation yet for the strange shakes that came before Redoubt’s eruption.

Volcanic eruptions are usually preceded by increased earthquake activity. Networks of seismic detectors have been erected around highly active volcanic regions like Alaska’s Aleutian Range to record these rumbling warning signs. In early 2009, Redoubt Volcano, a 10,000-foot peak southwest of Anchorage, began to shake and swell. In anticipation of an eruption, a team from the Alaska Volcano Observatory scaled the snowy slopes in order to deploy additional sensors.

Helena Buurman placing seismic sensorsHelena Buurman of the Alaska Volcano Observatory installs a seismometer on Redoubt Volcano in March 2009. Credit: Cyrus Read, Alaska Volcano Observatory/USGSOn March 21, just a day after the team from the observatory finished placing their sensors on Redoubt’s snowy slopes, the peak released an explosive cloud of ash, reaching heights of 18 kilometers above sea level and sending volcanic mudflows into the valley below. It was an impressive blast, but not out of the ordinary. Only later, when scientists from the University of Washington took a look at the data beamed back from those seismic sensors, did they notice something odd.

Alicia Hotovec-Ellis, a University of Washington graduate student studying volcano seismology, identified a series of rapid, repeating quakes in the moments just before the ash explosion. A similarly ordered pattern of tremors, neatly spaced like the teeth of a comb, had been seen before in the “drumbeat earthquakes” accompanying a minor eruption of Mount St. Helens between 2004 and 2006, but the Alaskan quakes were happening faster than anyone had ever seen.

Hotovec-Ellis published her findings early last year in the Journal of Volcanology and Geothermal Research. She converted some of her seismic data into sound and sped it up. The resulting half-minute clip offers a frightening prologue to an eruption.

Listen to what she calls a seismic scream below:

[audio mp3="http://more-deals.info/images_blogs/wiredscience/2013/07/volcano_1.mp3%22%5D%5B/audio%5D The Redoubt tremors followed a harmonic pattern, reminiscent of popping corn kernels, but Hotovec-Ellis noticed that their frequency grew and grew as the explosive eruption approached. While the St. Helens quakes had maxed out at a rate of five per second, the Redoubt quakes went off at six times that rate, eventually building into a frightening hum of harmonic noise.

More peculiar still, moments before the ash cloud began to spew, the “seismic screams” went quiet. What was causing this swelling storm and sudden silence?

This is the crucial question for John Vidale, Hotovec-Ellis’ adviser at the University of Washington. “There’s a lot of reasons why this could be happening,” he said. “We’re finally starting to narrow down these possibilities.” Previous theories ranged from a plug of rock creaking like a pressurized champagne cork to flute-like vibrations within the tubes of magma.

Fumarole on Redoubt VolcanoFumarole on Redoubt Volcano. One day before the eruption of March 22, 2009 the volcano shows steam venting and the summit glacier melting. Credit: Cyrus Read, Alaska Volcano Observatory/USGSVidale and Hotovec-Ellis called on Dunham’s team at Stanford to create a model to explain the geologic events that were causing the seismic scream. In Dunham’s new work, he was able to successfully reconstruct the dynamics of the harmonic earthquakes on a computer, and calculate the enormous pressures on the rocky faults within the mountain.

Huge rifts like California’s San Andreas Fault stretch hundreds of kilometers in length and spawn devastatingly powerful earthquakes. In contrast, the faults within Redoubt Volcano are on the scale of just meters, producing much weaker quakes that only wobble the ground by millionths of a meter in each direction. But that energy, Dunham says, is very focused.

According to his model, fault pressure builds up to more than 100 times higher than atmospheric pressure in the blink of an eye. Dunham explained it would be equivalent to diving down one kilometer below the sea in a single second. The repeating pops are probably caused by friction as the viscous magma slips past the rock around it, the molten fingernails on a rocky chalkboard. Eventually, in the moments just before the volcanic explosion, the friction gives way to gliding, an instant of calm before the storm.

Geophysicist Clifford Thurber of the University of Wisconsin-Madison, who was not involved in the research, thinks Dunham’s version of events makes sense, but advises caution. “It seems to be able to model the observed phenomenon,” he explained. “But any time you have a model with several parameters, you can often tweak it to get the data to fit.”

While this kind of research is important for understanding the geologic sequence of events leading to an eruption, Hotovec-Ellis says seismic screams occur too close to the big event to be useful for for volcano prediction. “By the time that we saw these things happening, the volcano was already erupting,” she said.

The ash plume from Redoubt Volcano rises nearly 65,000 feet above sea level on March 26, 2009, as seen from Japan’s MTSAT satellite. Credit: UW-Madison SSEC/CIMSS