Accessibility links

Breaking News

Researchers Forecast Volcanic Eruptions Using Satellite Data 

Mount Sinabung spews volcanic material during an eruption in Karo, North Sumatra, Indonesia, March 2, 2021.
Mount Sinabung spews volcanic material during an eruption in Karo, North Sumatra, Indonesia, March 2, 2021.

Scientists appear one step closer to predicting volcanic eruptions — a problem that has vexed volcanologists for decades. Research published last week in Nature Geoscience found that using satellite observations to calculate how quickly underground molten rock, or magma, accumulates beneath volcanoes could forecast certain eruptions weeks or months in advance.

"Any kind of information we can use to get at this forecasting thing is going to be important, because the more time you have to warn people that they can take some action, the more you can decrease the impacts of eruptions," volcanologist Michael Poland of the United States Geological Survey told VOA. "That's all we have, really, in terms of decreasing eruption impacts — to get out of the way."

Most volcanoes don't erupt without warning. They swell up, set off small earthquakes and let off gas leading up to an eruption — what volcanologists call "unrest." But while volcanoes rarely erupt completely out of the blue, it's also not uncommon for unrest to settle down without an eruption.

"The challenge is to understand when these changes in these monitoring parameters will lead to eruption, and when it doesn't," Federico Galetto, a volcanologist at Cornell University and first author of the new study, told VOA.

Currently, the gold standard for eruption forecasting involves highly localized observation of individual volcanoes, said Poland. But most volcanoes aren't closely monitored on the ground. In contrast, deformation — how volcanoes bulge and distort during unrest — can be measured from space for even the most remote volcanoes.

"The satellite deformation technique has really shown that a lot of these volcanoes inflate and deflate, and that allows us to help get to that sort of forecasting 'Holy Grail' in some places where there aren't ground-based data," said Poland.

Smoke billows from Mount Etna volcano, the largest of Italy's three active volcanoes, near the Sicilian town of Catania, southern Italy, July 20, 2019.
Smoke billows from Mount Etna volcano, the largest of Italy's three active volcanoes, near the Sicilian town of Catania, southern Italy, July 20, 2019.

Unfortunately, deformation alone can't reliably forecast eruptions. But Galetto and his colleagues thought that magma flow rate, which can be calculated using deformation data, might work better.

To find out, they considered 45 episodes of unrest in basaltic calderas — common volcanoes that usually look like flat, broad shields of dark basalt rock, including the volcanoes of Hawaii, Iceland and the Galápagos Islands. Basaltic calderas are considered relatively easy to study thanks to relatively shallow magma chambers — pools of molten rock beneath the Earth's surface — and frequent eruptions, and they have been observed for a long time.

"They picked a subset where we have a lot of information and a lot of observations, these basaltic calderas," said Poland. "These types of volcanoes, we have a lot of experience with … they tend to be great laboratories."

Galetto's analysis revealed that magma flow rate reliably predicted whether unrest would end in a magma chamber rupture — which usually causes eruption — or just fizzle out.

All volcanoes in the dataset with magma flow rates greater than one-tenth of a cubic kilometer per year — roughly 40,000 Olympic swimming pools — ruptured their magma chambers within a year. Inflow rates 10 times lower didn't lead to a magma chamber rupture in 89% of cases, and never before more than a year of unrest. Volcanoes with middling flow rates were harder to predict, with factors like rock type and magma chamber size coming into play.

"This is really promising," said Galetto. "That seems to [be] working very well in these types of volcanoes."

Calculations by Galetto and his team suggest that low magma flow rates don't tend to cause eruptions because slow-filling magma chambers behave a bit like viscous silly putty or molasses, oozing outward to accommodate a slow trickle of incoming magma without rupturing. Fast flow rates drive up pressure abruptly enough to crack magma chambers instead of just squeezing them.

"That makes sense," said Poland. "The faster you blow up the balloon, the more likely it's going to pop." But he also cautioned it's going to be a challenge to use the new results to forecast specific volcanoes.

"In volcanology, there's always a level of local expertise for your volcano that's needed, because every volcano is different," he said. "But we can learn some general trends … that can help us out in guiding us in the right direction when we are looking at these specific systems we're trying to forecast.

Based on his results, Galetto thinks magma flow rate could help forecast eruptions weeks or months ahead for basaltic calderas. But there's still work to be done. Fine-tuning forecast calculations with volcano-specific data as Poland described will be important for making good predictions, he said, as will collecting and analyzing better satellite deformation data.

"My paper is just a starting point, not the ending point," said Galetto. "We should start … to see if this relationship can be found out in other volcanoes. Because the other point is to try to extend these results not only to the group of volcanoes that I studied but also to try to extend these results to other groups of volcanoes. And it will be much more complicated."