As Erebus Lives and Breathes

MCMURDO STATION, ANTARCTICA — Even when the December sun beats down 24 hours a day, most of Antarctica remains cold, if not brutally frigid. With one dramatic exception. Wind-blown clouds of steam rise year-round from a lava lake atop Mount Erebus, the planet’s southernmost active volcano.

This ice-covered cone belongs to a small chain of otherwise dormant peaks that make up Ross Island. Some 1,300 summer residents at the National Science Foundation’s McMurdo research station 35 kilometers away can revel in the picture-postcard backdrop that Erebus offers. Few, however, have scaled the nearly 3,800-meter summit to peer into its churning pool of molten rock — a lake of lava roiling at roughly 1,000° Celsius.

“This lava lake is a window into the volcano and its magma chamber,” says volcanologist Philip Kyle of New Mexico Tech in Socorro. The magma chamber is the heart of a volcano; it controls and governs all eruptions. The lava in the lake rises up from inside a magma chamber somewhere deep below the primary crater on Erebus.

In most volcanoes, magma hides beneath a rocky cap. Only a few volcanoes, like Erebus, have lava lakes that have been open to the sky for decades, making them prime spots for the study of volcano behavior. And ironically, Kyle says, Erebus is perhaps the most accessible of such volcanoes — with the best logistical support thanks to transportation and other resources available through NSF at McMurdo. That’s why Kyle — one of the most seasoned of Erebus watchers — has trekked to Antarctica annually for four decades.

Along with shifting teams of other researchers, Kyle collects the volcano’s vital signs in an ongoing effort to gain clues to its inner workings — clues that might help scientists understand volcanoes elsewhere on the planet. With an ever evolving battery of physical and chemical tests, these Erebus observers have been probing the volcano’s dynamic plumbing system beneath the magma chamber. Its conditions affect whether Erebus just exhales streams of carbon-rich gas or snorts out the occasional car-sized lava bomb.

To understand it all, Kyle and his colleagues focus on water vapor and other gases that bubble up through the magma. Effectively Erebus’ blood gases, these volatile chemicals rocket up from below Earth’s crust. The bubbles’ size and content offer valuable clues to their origin — and the volcano’s stability.

Several years ago, Erebus watchers stumbled upon subtle cyclic variations in the gases emanating from the lava lake. The proportion of different gases changed in a pattern that repeated every 10 to 18 minutes. In sync with this oscillating gas chemistry, emerging data show, the lava lake’s surface experiences a rhythmic 2- to 3-meter rise and fall. “In simple terms,” Kyle concludes, “Erebus breathes. And it’s the first time we’ve been able to see the breathing of a volcano. By watching this, we get a greater understanding of how all volcanoes work. That’s the real bottom line.”

Listening to the inaudible

For scientists and the public alike, a big concern is whether an active volcano stands poised to begin a wholesale eruption or is merely suffering some geo-indigestion. So in 2000, Kyle invited one of his students, geophysicist Jeffrey Johnson, to set up a small network of microphones inside the rim to record the gastric churnings at Erebus.

Generated by the explosive release of gas at the lava lake’s surface, the primary rumblings center around 1 hertz, below the threshold of human hearing. Their intensity depends on how big the breaking bubbles are and the pressure inside them. Now Johnson’s microphones listen to the gurgling year-round, not just during that brief, relatively warm period when the summer sun shines day and night. And the infrasound ears aren’t diminished by the pea-soup fog that often shrouds the lake.

As his initial data poured in, Johnson says, he was amazed at the infrasound’s intensity. If audible, it would be booming at an ear-damaging 140 decibels, equivalent to gunshots or a jet engine at close range.

By knowing something about the depth and pressure at which the gas bubbles were born, Johnson began to back-calculate how big they had to be to trigger such booming infrasound. “We concluded we were recording the explosions of gas bubbles that were enormous — 20 to 40 meters in diameter!” he says.

Johnson says he can still recall Phil Kyle telling him that bubbles that big didn’t exist at Erebus. But a few years later, Kyle would have to apologize. A camera installed at Erebus to record nonstop video of the lake surface corroborated Johnson’s calculations about the bubbles’ extraordinary size.

Such megabubbles, known as slugs, probably fill the entire diameter of the conduit they ascend to reach the lake, Johnson says. During some particularly active periods — as in 2005 and 2006, when Erebus was regularly flinging lava bombs up and over the crater rim — such slugs explode throughout the day for weeks or months. During the volcano’s current relatively quiet phase, such gas slugs may break the lake’s surface only once a week, says Johnson, now at Boise State University in Idaho.

By analyzing the bubble sizes and the gurgling rate of the lava lake, he and others are homing in on sources of the eruptive gases, how fast their bubbles rise, how much the bubbles coalesce during ascent and what the magma conduits must look like. There are still plenty of unanswered questions, Johnson says, “but infrasound is helping us better understand these systems.”

Experimental detonations

Seismologists rely on a different set of ears to monitor Erebus.

Earthquakes have sometimes accompanied the volcano’s more dramatic eruptive events, particularly in the late 1980s. So a year-round seismic monitoring system listens for signs of rattling on or below the volcano. Some Erebus watchers have begun using that technology to map the volcano’s interior.

In 2007, Richard Aster and colleagues at New Mexico Tech “listened” to the seismic waves scattering through Erebus every time it erupted. Those eruptions tend to be mild and frequent — and generally consistent in size. “So we’ve got a repeating air gun–like seismic source right in the middle of this volcano,” Aster says, “that’s popping over and over.”

Those data hinted at the location of the volcano’s magma reservoir — below and to one side of the lava lake — and to other hot spots dotting the volcano’s upper interior.

Daria Zandomeneghi, now a fellow at the Abdus Salam International Centre for Theoretical Physics in Trieste, Italy, led a related Erebus mapping project while working with Kyle at New Mexico Tech. Her team planted explosive charges in 20-centimeter diameter ice holes at depths of 7 to 15 meters. A network of portable seismic sensors then recorded the shock waves generated by a sequence of 12 detonations.

“You might say we’ve been trying to CT scan Erebus by throwing 100 seismometers on top of it and then running around setting off explosions,” says Kyle.

Unlike a medical CT scan, which uses X-rays, volcanic CT scanning analyzes how quickly vibrations zing through the mountain. Those passing through cool rock travel speedily; others slow as they encounter hot rock or conduits containing liquid magma. By knowing when the detonations occurred, the distance the ground-shaking waves traveled and the time it took them to reach each seismometer, the team could map regions of varying temperature. By combining that map with the seismic data from the volcano’s eruptions, says Kyle, “We’ve been able to map the innards of Erebus.”

But only roughly.

Merging the data identified just one common hot spot — probably the magma chamber — 500 meters below the surface and about 500 meters northwest of the lake, Zandomeneghi says. But the volcanic CT scans cannot establish the magma volume or identify the apparently small conduits between chamber and lake. “A feeding system obviously exists for the lava lake,” she says. Her team’s data indicate that it must consist of one or more narrow-diameter shafts.

The complicated structure of the system feeding the lake might explain how slight changes in the activation of feeder shafts can affect the frequency and location of eruptions, says Zandomeneghi.

The latest scanning findings will appear soon in the Journal of Geophysical Research.

An emerging portrait

Setting off explosions isn’t necessary to glean clues to Erebus’ source. A bit of sniffing will do, and Erebus’ breath is unique. The dominant gas exhaled by most volcanoes is water vapor, Kyle notes, typically accounting for 60 to 90 percent of the total.

Carbon dioxide, the next biggest gas constituent, usually makes up 5 to 15 percent at volcanoes around the world. But Erebus emits a surprisingly high proportion of CO₂. Recent measurements now indicate that 90 percent of the volcano’s normal gas output consists of a roughly equal mix of water vapor and CO₂, although the CO₂ varies a bit with time and how actively the volcano is erupting.

Another surprise: Carbon monoxide constitutes the third most prevalent gas in the volcano’s releases. There’s always a little CO in volcanic gases. But two years ago Kyle’s team discovered unusually high amounts at Erebus.

All of this carbon may help explain why Erebus even exists, says geochemist Erik Hauri of the Carnegie Institution for Science in Washington, D.C. Volcanoes emerge when molten rock rises from below the Earth’s crust. They tend to develop in places where tectonic plates collide or where one plate thins and fractures. But because copious carbon lowers the melting point of mantle rocks, Hauri explains, it allows regions of the deep Earth to melt when they otherwise wouldn’t. “So you have these volcanoes, like Erebus, that sort of pop up in the middle of a plate,” Hauri notes.

Carbon compounds are also among the least soluble gases in magma, so they will tend to bubble out. Those bubbles can enhance a magma’s buoyancy. The carbon-rich gases at Erebus might even help explain the volcano’s puzzling low-level eruptive behavior, Hauri says. “It’s kind of cooking all of the time.”

That simmering is visible in the lake, where a roughly 30-meter-diameter stew of molten rock studded with crystals of the mineral feldspar churns away. The lake has been coughing out lava bombs containing these crystals for centuries. In his office at McMurdo, Kyle keeps a huge stash (above). He views them as mineral fingerprints of the maturation of the volcano’s magma.

Although many magmas host microcrystals, Kyle notes, the crystals at Erebus are of phenomenal size, as long as 8.3 centimeters.

The feldspar and its glass inclusions form at different times and places during the magma’s ascent. So analyzing the crystals’ makeup provides further information about how the magma coursing through Erebus has been evolving.

Although magmas from many volcanoes erupt looking pretty much like a melted version of the original source rock, some magmas — like at Erebus — are more evolved. In fact, as volcanoes go, Erebus is fairly high on the geo-evolutionary ladder, observes volcanologist Clive Oppenheimer of the University of Cambridge in England.

As a volcano forms, high-pressure magma deep underground forces its way upward, making Earth’s surface bulge. Eventually a stream of lava erupts out, its chemical recipe almost identical to that of the original magma. Over time, though, twists and turns can develop in the conduits channeling the magma upward. At Erebus, that process probably creates pockets that cause the magma to get temporarily hung up as it rises. During these pauses, certain gases may bubble out; others may catch up with the magma to enrich their concentration inside it. And various constituents of the magma may interact chemically during delays in its ascent.

So the initial magma “may get stewed and brewed in the lower crust beneath the volcano,” Oppenheimer says. By the time it erupts at Erebus, “you might have left 75 percent of the starting material behind. From that perspective, we’d call this magma quite evolved.”

The type at Erebus is known as “phonolite.” To understand how it morphed from a primitive liquid rock into this evolved state, Oppenheimer’s team has been applying some forensic geochemistry to samples of the erupted rock and to gases bubbling from the lava lake.

The CO₂ measured by Oppenheimer and colleagues in belched gases indicates that Erebus is pumping carbon from more than 16 kilometers down, well within the mantle. Such deeply derived CO₂ will extract water from the molten rock. “This drying out is one way that magmas evolve,” Oppenheimer explains.

This water extraction helps portions of the magma to crystallize, he says, trapping tiny quantities of molten rock, much as resin can trap insects in what will harden as amber. The crystal’s trapped glassy bits amount to “fossilized bits of the magma chamber,” Oppenheimer says. “We can see how much CO₂ and how much water is in those tiny inclusions and tell how deep the sample came from.”

The Antarctic trade-off

There’s no way to get around the weather in Antarctica. The field season at Erebus runs just the month of December, maybe a few weeks longer. Anyone who arrives much earlier or stays much later faces beastly cold and snowy conditions. And for half a year or more, there is too little light to fuel solar-powered equipment. It’s frustrating, Oppenheimer notes, to realize how much energy is being radiated from the lava lake — “and we’re not able to use any of it.” And howling winds for months on end are brutal on unattended gear.

But there are pluses: It’s easier to measure water vapor, crucial information for scientists, here than it is anywhere else in the world. “At many volcanoes water vapor represents more than 90 percent by weight of the gases expelled,” notes Alain Burgisser at the University of Savoy in Bourget-du-Lac, France. But atmospheric moisture can mask the puffs of water vapor seeping from a volcano — except at Erebus, where the air is nearly bone dry.

The lava lake at Erebus also provides “remarkably fresh, clean sources” of vented gases and vapors that, at other sites, may have taken a torturous path through rock or water baths, morphing chemically or becoming diluted along the way. And Erebus’ eruptions are predictably regular, another advantage for scientists. A lake at Hawaii’s Kilauea volcano may be more thoroughly studied, but its varying eruption rates and intensity make understanding its ongoing activity more challenging.

Even Erebus’ isolation offers benefits. Human settlements — in some cases major metropolitan areas — have developed in the foothills of many volcanoes around the world. No one would dare shake and rattle those peaks for fear of putting whole communities at risk, Kyle says. And he points out that civil unrest has postponed research at two other volcanoes with especially long-lived lava lakes — Erta Ale in Ethiopia and Mount Nyiragongo in the Democratic Republic of the Congo. At Erebus, Kyle says, scientists can work in relative safety, learning things that might translate to many other volcanoes.Erebus is an “archetype volcano,” says Oppenheimer — even if the window to its heart is easily accessible for only about a month each year. Ongoing study here is likely to increase understanding of what can be one of Earth’s most powerful and unpredictable geologic events.

Alive on Erebus?

Carsten Peter/National Geographic Stock

Not all of the gases emanating from Erebus bubble out of its lava lake. The same magma chamber that feeds the lake also sends shafts of scalding magma up near the surface at various spots around the volcano’s summit. At many sites, hot gas breaks through or melting ice can percolate down to hot rock before flashing to steam. In either case, the resulting hot gas can tunnel out room-sized enclosures beneath the ice (above). Among Antarctica’s more magical — if ephemeral — environments, these warren-like ice caves may host life that dines on volcanic rock, scientists now suspect.

Any organisms would be microbial and very slow-living, says geochemist Hubert Staudigel (below) of the Scripps Institution of Oceanography in La Jolla, Calif. “When we look at volcanic glass — the stuff that forms when lava doesn’t have enough time to cool and form crystals — we find micron-sized drill holes.” Running up to 100 micrometers in length, he notes, these resemble holes that bacteria or fungi carve into rock at deep-sea vents and other mineral-rich sites.

To test the idea that germs ate the holes into the Erebus glass, Staudigel and his colleagues put out several polished mineral samples as bait, beginning four years ago. Preliminary data collected from some of the bait samples two years ago turned up signs of unidentified microbes, he says. While promising, that didn’t prove the germs had come to dine.

This past December Staudigel’s group retrieved additional bait samples. In coming months they will begin probing for signs that microbes have begun lunching on them.

Roberto Anitori/Oregon Health and Sciences Univ.