Hidden patterns in seismic signals could help forecast eruptions and prevent loss of life on volcanoes, according to pioneering University of Canterbury research spurred by the deadly Whakaari (White Island) eruption.
University of Canterbury (UC) Civil and Natural Resources Engineering postdoctoral fellow Dr Alberto Ardid has used machine learning to identify a seismic frequency pattern that occurred in the days leading up to the deadly Whakaari eruption on December 9, 2019.
Dr Ardid, a Chilean geophysicist, studied recordings from GeoNet seismometers – an instrument that measures ground noises and shaking – prior to 18 eruptions across six active volcanoes around the world. This included three in New Zealand; Ruapehu, Tongariro, and Whakaari in the Bay of Plenty.
A machine-learning algorithm allowed Dr Ardid to sift through thousands of recordings and highlight particular frequency patterns that occurred regularly before an eruption. His findings are published in the international journal Nature Communications today. [9pm NZT, April 20, 2022]
“We discovered a change in frequency that tells us an eruption is much more likely to occur over the next few days,” Dr Ardid says. “These frequency transitions allow us to listen in on what’s happening at different depths under the volcano. Watching closely for this pattern could provide an early warning of future eruptions.”
The key discovery was a peak in the displacement seismic amplitude ratio (DSAR) which suggests a blockage in the shallow part of a volcano. It signals that a seal or lid has formed, which traps hot gas, builds pressure, and sometimes triggers an explosion.
“We found the same DSAR signal would build and peak in the days before the last five eruptions at Whakaari, and many of the eruptions at Ruapehu and Tongariro,” Dr Ardid says.
He believes the discovery is significant because the warning signal from one volcano can, under the right circumstances, be transferred to other volcanoes that don’t have long histories of recorded eruptions.
“What we found is that this precursor to an eruption is common among several volcanoes, particularly the New Zealand ones which have a hydrothermal system beneath the crater. The next step is to see how well this signal works as a warning system for volcanoes around the world.
“It’s really exciting that we can potentially contribute to a warning system that could help save lives.”
Dr David Dempsey, a Senior Lecturer in Civil and Natural Resources Engineering at the University of Canterbury who worked closely with Dr Ardid on the research and is a co-author on the journal article, says the DSAR signal could become part of a precautionary evacuation alarm system, particularly in tourist areas.
“Active volcanoes, including Whakaari, Ruapehu, Tongariro, and others around the world where visitors and skiers are likely to be nearby, are unpredictable and sometimes hazardous. Early warning systems could save lives and avoid debilitating injuries.
“This research was motivated by the tragedy on Whakaari, but we think it has wider application because there are several other New Zealand volcanoes that have a similar style of eruption.
“We would like to use this for real-time hazard information, perhaps to let tourists know when it’s safer to visit volcanoes.”
Dr Dempsey says lessons learned from Whakaari and Ruapehu could be applied to other volcanoes that haven’t erupted yet, or don’t have large monitoring networks. However, there would need to be monitoring for the warning signal, scientists to interpret the data, and an organisation that decides when to issue alerts or restrict access to a volcano.
Dr Ardid’s research was co-written with Dr Dempsey, Professor Shane Cronin (University of Auckland) and Assistant Professor Corentin Caudron (University of Brussels, Belgium) who first developed the DSAR signal.
IMAGE CREDIT: University of Canterbury, University of Auckland, University of Brussels, Belgium