Popocatépetl, a volcano near Mexico City, may warn of its eruptions in its gassy belches.

Atonaltzin/iStockphoto

Volcanic fumes warn of imminent eruptions

Last month, researchers from the National Autonomous University of Mexico hoped to reach the top of Popocatépetl, a 5400-meter-tall volcano near Mexico City, to install monitoring equipment at its summit crater. But El Popo, as locals call it, rebuffed them with ash and belches of acrid gas—precisely what the scientists wanted to measure. They settled for installing the sensor lower down the mountain, and hope to move it higher next year. The goal is to measure just what—and how much—El Popo has been smoking, because the fumes may hold a promising way to forecast eruptions.

A growing body of monitoring data suggests that a sharp jump in the ratio of carbon to sulfur gases emanating from a volcano can provide days to weeks of warning before an impending outburst. The latest evidence comes from three recent studies, focusing on volcanoes monitored as part of the Volcano Deep Earth Carbon Degassing (DECADE) initiative. They offer hope that geochemical monitoring of gases could someday join the two geophysical mainstays of forecasting: tracking the swelling of Earth’s surface and the rise in earthquakes that typically precede eruptions. “It’s statistically robust as a forecasting tool,” says Tobias Fischer, a volcanologist at the University of New Mexico in Albuquerque, and chair of the DECADE project.

The idea of sniffing out restlessness in volcanic fumes has been around for decades. For instance, a sharp rise in sulfur emissions helped scientists anticipate the 1991 eruption of Mount Pinatubo in the Philippines. Scientists have also keyed in on the carbon-to-sulfur (C-S) ratio in volcanic gases as a particularly helpful metric. In principle, it can signal when a fresh injection of magma is rising from deep in the crust—a prelude to an eruption.

The ratio changes because carbon dioxide (CO2) dissolved in rising magma bubbles out at depths of 10 kilometers or more, as the pressure drops. Sulfur-rich gases, in contrast, stay in solution up to shallower depths. A spike in the ratio can thus provide warning that a new batch of magma has risen above a deep threshold. A subsequent drop in the C-S ratio could indicate that the magma has climbed further, to depths where sulfur gases are released, but Fischer says this hasn’t been observed enough to be reliable.

A whiff of future eruptions

Spikes in gas ratios occur just days before eruptions at Turrialba, a volcano in Costa Rica.

(Graphic) J. You/Science; (Data) Turmoil at Turrialba volcano (Costa Rica), Maarten de Moor et al., 2016

Despite the simple mechanism, establishing a clear link between the ratios and eruptions requires constant monitoring. Historically, researchers just bottled a few gas samples during a visit to a volcano or used airplanes or remote-sensing tools to watch a volcano for several days or weeks, says Christoph Kern, a physicist with the U.S. Geological Survey in Vancouver, Washington. Either way, Kern says, it was hard to catch an eruption in the act.

But that changed in the early 2000s, when scientists began to develop new devices that could be left on volcanoes to make continuous measurements and transmit the data to researchers. They were solar powered, hardy enough to survive the elements, and cheap enough to risk sacrificing in an eruption. “They’re essentially expendable,” says Marie Edmonds, a volcanologist at the University of Cambridge in the United Kingdom.

Italian scientists were the first to deploy these instruments at volcanoes like Etna and Stromboli, and they began to notice changes in the C-S ratio in the days and hours prior to eruptions. Since then, U.S. and Japanese geologists have installed instruments at a handful of volcanoes in those countries, and the DECADE project has added them at nine more around the world, including El Popo. Overall, changes in C-S gas ratios seem to be a powerful portent, Fischer says. “Now, we’re seeing it at many different volcanoes.”

Perhaps the clearest illustration comes from Turrialba in Costa Rica, a volcano that poses a threat to the city of San José, 30 kilometers to the west. Maarten de Moor, a researcher at the Volcanic and Seismic Observatory of Costa Rica, helped install gas sensors on Turrialba in early 2014, just in time for the volcano to start erupting. He led a study, published in the Journal of Geophysical Research in August, reporting sharp increases in the C-S ratio of gases a few weeks before each outburst over two eruption cycles (see chart, above). “What we’ve seen is quite mind-blowing,” he says. “These signals are eye-opening.”

But for monitoring gas ratios to become a widely used forecasting tool, researchers will need to understand many complicating factors, says Clive Oppenheimer, a volcanologist at the University of Cambridge. “The interpretation of gas chemistry, particularly for the purposes of forecasting, is not an exact science,” he says. “Very far from it.”

At Turrialba, for instance, there were different sulfur gases in the mix. Sulfur dioxide (SO2) gas from the magma interacted with underground water to produce hydrogen sulfide during the first eruptive episode, but not the second. De Moor says these observations could indicate that the water eventually boiled off, or that new volcanic conduits formed, bypassing the water reservoirs. At Poás, another Costa Rican volcano, the summit crater contains an acid lake that normally absorbs the SO2 percolating through it but allows the CO2 to pass through unimpeded—keeping the C-S ratio relatively high even when an eruption isn’t imminent. But DECADE’s monitoring efforts have revealed that, in the days before an eruption at Poás, the emissions of sulfur gases spike, exceeding the lake’s ability to scrub out the sulfur and causing the C-S ratio to plummet. It’s the opposite signal from the one seen at places like Etna and Turrialba, but it’s equally reliable, Fischer says. 

Satellites could theoretically help researchers monitor many of the world’s 550 historically active volcanoes from orbit. Instruments aboard NASA’s Terra satellite, for instance, can already measure volcanic sulfur emissions reasonably well. But researchers are still working to measure SO2 and CO2 at the same time, and measuring point sources of CO2 is challenging because of high background levels in the atmosphere. Even a big CO2 burp from a volcano only increases the concentration measured by satellites by less than a percent, says Florian Schwandner, a geochemist at NASA’s Jet Propulsion Lab in Pasadena, California. 

For now, the scientists who want to explore the forecasting power of the C-S ratio must wait for ground-based monitors to capture more eruptions. And maintaining these sensors can be a hassle, De Moor says. Even small dustings of ash can cover up solar panels or damage electronics. That’s what caused a sensor on Turrialba to stop transmitting data in May, forcing De Moor to visit once a week to download it in person—sometimes in dangerous conditions. But he says he’s always careful, and tries to remember a bit of wisdom passed down from Fischer, his Ph.D. supervisor, about taking risks in the name of science. “You are going to make more contributions if you actually survive this.”

*Correction, 23 November, 4:17 p.m.: A previous version of the story incorrectly stated that the Carnegie Institutions for Science manages DECADE. In fact, DECADE is a program run by the Deep Carbon Observatory, an international effort led by Carnegie.