It wasn't what they were looking for but that only made the discovery all the more exciting. In January 2010 a team of scientists had set up two crossing lines of seismographs（地震仪） across Marie Byrd Land in West Antarctica. It was the first time the scientists had deployed many instruments in the interior of the continent that could operate year-round even in the coldest parts of Antarctica.

Like a giant CT machine, the seismograph array used disturbances created by distant earthquakes to make images of the ice and rock deep within West Antarctica.

There were big questions to be asked and answered. The goal, says Doug Wiens, professor of earth and planetary science at Washington University in St. Louis and one of the project's principle investigators, was essentially to weigh the ice sheet to help reconstruct Antarctica's climate history. But to do this accurately the scientists had to know how Earth's mantle would respond to an ice burden, and that depended on whether it was hot and fluid or cool and viscous（粘性的）. The seismic data would allow them to map the mantle's properties.

In the meantime, automated-event-detection software was put to work to comb the data for anything unusual.

When it found two bursts of seismic events between January 2010 and March 2011, Wiens' PhD student Amanda Lough looked more closely to see what was rattling the continent's bones.

Was it rock grinding on rock, ice groaning over ice, or, perhaps, hot gases and liquid rock forcing their way through cracks in a volcanic complex?

Uncertain at first, the more Lough and her colleagues looked, the more convinced they became that a new volcano was forming a kilometer beneath the ice.

The discovery of the new as yet unnamed volcano is announced in the Nov. 17 advanced online issue of Nature Geoscience.

Following the trail of clues

The teams that install seismographs in Antarctica are given first crack at the data. Lough had done her bit as part of the WUSTL team, traveling to East Antarctica three times to install or remove stations in East Antarctica.

In 2010 many of the instruments were moved to West Antarctica and Wiens asked Lough to look at the seismic data coming in, the first large-scale dataset from this part of the continent.

"I started seeing events that kept occurring at the same location, which was odd, "Lough said. "Then I realized they were close to some mountains-but not right on top of them."

"My first thought was, 'Okay, maybe its just coincidence.' But then I looked more closely and realized that the mountains were actually volcanoes and there was an age progression to the range. The volcanoes closest to the seismic events were the youngest ones."

The events were weak and very low frequency, which strongly suggested they weren't tectonic in origin. While low-magnitude seismic events of tectonic origin typically have frequencies of 10 to 20 cycles per second, this shaking was dominated by frequencies of 2 to 4 cycles per second.