How Satellite Imagery Can Help Predict Volcanic Eruptions
Recently, the Ubinas volcano erupted in Peru, leading to the evacuation of thousands due to explosions and ash. And earlier this month, the Stromboli eruption killed one hiker and caused vegetation to catch fire. Both instances are a sobering reminder that volcanic activity can wreak havoc on surrounding areas. But while it’s impossible to prevent eruptions, we may be able to better predict volcanic eruptions, prepare for them and reduce impending destruction.
High resolution satellite imagery can be fundamental to identifying volcanoes that are at risk of blowing their tops, experts say, and should be included in volcano monitoring in the future.
“I think constellations such as Planet’s fleet have a major role to play in the future of volcano monitoring from space,” says volcanologist Simon Carn, professor of geological and mining engineering sciences at Michigan Tech. “Planet’s high-resolution, high-cadence visible imagery might help us to identify potential precursors to eruptions … which could potentially provide time for populations and aviation at risk to be warned—and, if necessary—evacuated or diverted in advance.”
Precursors to volcanic eruptions could include color changes in crater lakes, damage to vegetation, and ice melting on ice-capped volcanoes, Carn says. These changes might happen over periods ranging from a few hours to years. Using Planet imagery, scientists can monitor these changes leading up to a possible eruption and assess their significance.
Planet’s data can also be useful during and after an eruption to map the extent of volcanic ash deposits, Carn explains, which can impact health and contaminate water supplies. Mapping the extent of deposits provides an estimate of eruption size, which is not always easy to measure.
Planet data can also give civil protection agencies information to determine which areas are most impacted. In the aftermath of an eruption, volcanologists can use the imagery to see changes that may have occurred on the volcano (like crater size and location) that could impact future volcanic activity.
There are roughly 1,500 active or potentially active volcanoes on Earth. Only about 50 percent are monitored from the ground, and only 35 percent of the volcanoes that have erupted since 1500 CE are continuously monitored by ground sensors, according to Carn. This is due in part to the remote locations of volcanoes and the high costs for ground sensor installation and maintenance.
While other kinds of satellites are indispensable—having been specially made to measure things like gas emissions, heat flux and ground deformation—Planet’s imagery is particularly valuable. This is because humans are more “tuned in to visible wavelengths,” Carn says, meaning we’re likely more adept at spotting visual changes—which may be why we’ve noticed more volcanic activity in recent years.
“In some ways, satellites are partly responsible for perceived ‘increases’ in volcanic activity on Earth, since more activity is now detected by the ever-increasing number of satellites in orbit,” Carn says. “But volcanism is the result of geological processes that operate on very slow, geological timescales—not human timescales. Random geological events such as volcanic eruptions will occasionally cluster in time and give the impression that activity is increasing, but there is no long-term trend.”
Also, Carn adds, “Volcanoes are not responsible for climate change,” despite the rumors to the contrary. Research has shown that volcanoes release less than one percent of the carbon dioxide released by current human activity. To put it into perspective, both land and undersea volcanoes create roughly 200 million tons of carbon dioxide every year, according to the U.S. Geological Survey, while industrial and automotive activities release 24 billion tons annually.
“It is still widely believed that volcanic carbon dioxide emissions exceed carbon dioxide emissions generated by people,” Carn says. “But this is not true.”
While satellites have been used to monitor volcanoes during and after eruptions over the last several decades, monitoring volcanoes before they explode is a new phenomenon—mostly because satellite sensor technology and sensitivity has greatly improved over the last 10 years.
A big challenge moving forward, Carn says, will be processing and analyzing daily collections of data for the Earth’s 1500-ish potentially active volcanoes. This may require the assistance of artificial intelligence and machine learning algorithms designed to identify such changes.
“One goal of volcanologists is to forecast the timing and size of eruptions,” Carn says. “The more data we have, the better the chances of doing this.”