Antarctic radar survey reveals a subglacial volcanic eruption

Like every other continent, Antarctica has hosted volcanic activity. Mount Erebus, the world’s southernmost active volcano, erupts continuously but nonthreateningly. Several volcanoes protruding from the West Antarctic ice sheet have produced bigger eruptions, the most recent of which was 7500 years ago. In 1993 a team led by Donald Blankenship of the University of Texas at Austin inferred from geophysical data ¹ that there is an active volcano beneath the WAIS, but they found no direct evidence of any specific eruption. Now Hugh Corr and David Vaughan of the British Antarctic Survey have found evidence of another subglacial volcano, also under the WAIS, and of its most recent eruption some 2300 years ago—probably the biggest eruption in Antarctica in the past 10 000 years. ²

If you think of Antarctica as shaped like an open fan, West Antarctica is the handle. Separated by the Transantarctic Mountains from the much larger and more stable East Antarctic ice sheet, the WAIS is a marine ice sheet: Parts of the bedrock to which it is anchored are far below sea level. If the entire WAIS melted, ocean levels would rise by 6 m. But a single volcano does not—and cannot—melt the ice sheet by itself. A cubic meter of molten rock can only melt 14 cubic meters of ice, and eruptions that produce more than 1 km³ of material are extremely rare. Corr and Vaughan’s eruption pierced a hole through the ice, through which the volcano could send a plume high into the sky. The volcanic ash and rock—collectively known as tephra—drifted tens of kilometers outward from the volcano and settled in a layer on the ice surface. Over the years more snow and ice accumulated on top of the tephra, trapping it within the ice sheet for Corr, Vaughan, and their colleagues to discover.

They found the tephra not by drilling and extracting an ice core—a process too expensive and time-consuming to carry out in more than a few carefully selected spots—but from radar data from an airplane flying overhead, a much faster and more efficient method for mapping the ice sheet. Sensitivity to frequency and phase allows the radar to probe not only the topography of the underlying bedrock and other opaque material but also structural details of the ice.

In the Southern Hemisphere summer of 2004–05, a team of US and UK researchers set out to survey the changing West Antarctic glaciers. The UK group’s focus was Pine Island Glacier, an ice stream that covers 10% of the WAIS. (There are, of course, no pine trees in Antarctica. Pine Island Glacier and the adjacent Pine Island Bay were named after the USS Pine Island, a ship that explored the area in the 1940s; the ship was named for Pine Island off the southwest coast of Florida.)

The UK group’s entire survey, sampling an area the size of Wisconsin, was meticulously planned in advance. Although Corr and Vaughan first noticed the ashen layer in the preliminary quality check of their data, they couldn’t go back and look at it more closely without taking time away from other equally important regions. Still, the data included several passes over the tephra, which revealed itself as a bright reflection distinct from the underlying bed, as shown in the figure.

Reliably determining the tephra’s volume, a measure of the eruption’s intensity, proved a challenge. Corr and Vaughan could estimate the layer’s thickness from the strength of the radar reflection. But because the tephra in the center was made up of larger rocks, not fine-grained ash, the reflection strength there varied so much that the lower and upper bounds on the volume, 0.019 km³ and 0.31 km³, differ by a factor of 16. However, volcanic eruptions vary by several orders of magnitude in the volume of ejected material, so even that wide range sets the eruption’s so-called volcanic explosivity index at either 3, for “severe,” or 4, for “cataclysmic.” (The 1902 eruption of Mount Pelée on Martinique in the Caribbean, the deadliest of the 20th century, had a VEI of 4. The 1980 eruption of Mount Saint Helens had a VEI of 5. Mount Erebus’s eruptions have VEIs of at most 2.)

From their radar data, Corr and Vaughan estimated the age of the tephra layer at three locations, including the one labeled X in the figure. They found that the tephra dates from 207 BCE, plus or minus 240 years. Ice cores provided a more precise date: The volcano emitted not only ash and rock, which settled nearby, but also hydrogen chloride, sulfur dioxide, and other gases, which traveled greater distances and left their mark on ice much farther away. Two ice cores, 1100 km and 1300 km away from the volcano, revealed that there was just one strong eruption in the time window inferred from the radar data. That eruption, the biggest in the region in the past 10 000 years, happened between 325 and 315 BCE.

Although the volcano hasn’t erupted for 2300 years, it could nevertheless be heating and melting the ice from underneath. A layer of water between glacier and bedrock would speed the process by which new snow pushes old ice aside and into the sea—resulting in a thinner glacier and a higher sea level (see Physics Today, March 2002, page 17 ). Pine Island Glacier has been flowing faster in recent decades, but there’s no reason to believe that the volcano has gotten any hotter over that time period. Even if it has, it could not have had any effect on other West Antarctic glaciers, which have also been thinning. The warming of the ocean waters remains the most plausible explanation for those changes, which are causing sea level to rise at a rate of 0.2 mm per year. Still, volcanic heat needs to be included in numerical models of ice flow. As Corr explains, “It is those models that will be needed to make predictions and influence policy.”