When volcanologists forecast an eruption, they can only offer different grades of warning from very low to very high threat levels. Unlike meteorologists, they can’t forecast the timing of the eruption, how many inches of ash will fall, or how long it will erupt. And in many parts of the world, without the resources to monitor volcanoes, people receive no warning at all.
Terry Plank (Lamont Doherty Earth Observatory, USA), a member of the Reservoirs and Fluxes Community and the DCO Executive Committee, thinks that with real-time multi-sensor data and physics-based models, volcanologists may one day be able to forecast volcanic eruptions just like meteorologists forecast hurricanes. Plank and her colleagues at the Lamont Doherty Earth Observatory, including co-lead Einat Lev, have received a five-year, almost $2.5 million grant from the Gordon and Betty Moore Foundation for their project, AVERT: Anticipating Volcanic Eruptions in Real Time. The researchers will monitor two volcanoes in Alaska in real-time to identify early warning signals and to develop an array of instruments that could be deployed worldwide for better eruption forecasting.
“Think about having to forecast a hurricane and all you have is one ocean buoy sitting out in the ocean and you have to go out there and collect the data,” said Plank. “That’s kind of where volcano science is.”
The project grew out of her work with the DCO and the National Academy of Sciences ERUPT report. “The DCO has done a lot for volcano forecasting, especially with the DECADE project,” said Plank. DECADE, which stands for Deep Earth Carbon Degassing, is an initiative to install carbon dioxide monitoring devices on 20 of Earth’s most actively degassing volcanoes. “DCO developed these real-time data streams that have captured new carbon precursors and may be revolutionary in anticipating eruptions.”
The AVERT team will install an array of instruments on two of the most active volcanoes in the U.S., Cleveland and Okmok in the Aleutian Islands. Cleveland produces small eruptions and is uninhabited, but more than 80,000 people fly over Alaska volcano air space each year. Okmok has had large eruptions in the past and is home to a small cattle ranch. Plank sees these neighboring sites as “good laboratory volcanoes” that are better representations of typical, explosive volcanoes, compared to ones that scientists have traditionally studied, like Hawaii and Etna, which are accessible and have frequent, low-energy eruptions.
The instrument arrays will include volcanic gas sensors, seismometers to detect micro-seismicity, GPS to monitor magma inflation in the ground, and infrasound detectors that listen for rising lava. Engineers at the Lamont Doherty Earth Observatory will develop satellite communications systems to transmit the data collected by the instruments in real-time. Thanks to recent advances in the bandwidth available through new satellite constellations, the data can be streamed, and will be openly available to the larger volcano research community.
DCO Reservoirs and Fluxes Community members Marten de Moor (National University of Costa Rica), Emma Liu (University College London, UK), and Kieran Wood (University of Bristol, UK) will contribute gas monitoring expertise. Liu will be using drones to drop “dragon eggs” – small, sacrificial sensors – into the crater, because the crater is too dangerous to access even from a helicopter. The AVERT team also is open to supporting other collaborators who have novel instruments to deploy.
“This array will be the first of its kind and we’d like to develop it so that it’s cheap and robust and we could move it to other volcanoes,” said Plank. When Guatemala’s Fuego volcano erupted with no warning in 2018, potentially killing more than 1,000 people, Plank realized that remote, real-time monitoring could forecast eruptions and potentially save lives in countries that don’t have the resources to maintain their own monitoring systems.
Previous DCO-supported research revealed that some volcanoes release a burst of carbon dioxide just before an eruption, but there are likely other clues that could inform volcano forecasting models. “We still have a lot to learn about the timing of events and from what depth the magma rises,” said Plank. “We really need data first, and so that’s the piece that we’re working on.”
Main image: Cleveland Volcano, one of the most active volcanoes in the world, sits in the Islands of Four Mountains group of the Aleutian chain. In this image of its summit vent, Carlile Island volcano lies in the background. Credit: Max Kaufman/Alaska Volcano Observatory/University of Alaska Fairbanks, Geophysical Institute