An alert system that could have given 16 hours’ warning of last year’s eruption at Whakaari/White Island is ready for
deployment, University of Auckland scientists say, with warning systems for Ruapehu and Tongariro the next priority.New Zealand does not currently have an advanced real-time warning system for volcanic eruptions. Instead, in line with
international best practice, GNS Science operate a Volcano Alert Level (VAL) system, which provides only a measure of
the current status of a volcano and is updated, generally every few weeks or months. It relies on human judgement and
consensus amongst scientists to spot activity that could signal a pending eruption.
Three weeks before last year’s fatal eruption at Whakaari, the VAL was lifted to Level 2, indicating heightened unrest.
After the eruption occurred, it was raised to Level 4 as per the design of the VAL system.
Dr David Dempsey from the University of Auckland says the eruption was preceded by a strong burst of seismic energy 17
“We think this was a sign that fresh magmatic fluid was rising up and pressurising water trapped in shallow rock and
loose deposits filling the vent.
“The resulting explosion was like a pressure cooker blasting its lid off. The early seismic burst is the most common
indication of imminent eruption at Whakaari.
“It’s a warning sign that could have been detected almost instantly by the forecasting system we have developed.”
With funding from the Ministry of Business, Innovation and Employment, the new system developed by Dr Dempsey and Dr
Andreas Kempa-Liehr from the Faculty of Engineering, and Professor Shane Cronin from the Faculty of Science, uses
sophisticated machine learning algorithms to ‘teach itself’ from the data fed into it.
It ‘learns’ from patterns in that data so that it is able to signal almost instantly when a particular pattern matches
that of the build-up to a previous eruption. With data from the past ten years at Whakaari the new system predicted four
out of five past eruptions missing only one unusual event – in 2016 - - which showed a different data pattern.
“Machine learning means it learns from ‘experience’ and so constantly improves its accuracy,” Dr Kempa-Liehr says.
A prototype of the forecaster has been operating continuously for five months and the development team are working with
GNS to implement it alongside their systems.
One of the challenges of the project was refining the threshold at which an alert is triggered. Currently the new system
uses a threshold of 8.5 percent probability an eruption is imminent, which, in simple terms, means it raises an alert
when there is a 1 in 12 chance of an eruption occurring. Each alert lasts about 5 days.
The researchers say that will mean a trade-off if the new system is adopted because the Island would likely be
off-limits for about one month a year.
“This system detects eruption types that are most likely to be fatal,” says Professor Cronin.
“The loss of life at Whakaari was a dreadful human tragedy and we hope that an automated forecasting system such as this
could avoid loss of life but it’s a question of whether policy makers, land owners, iwi and tourist operators want a
system that, once an alert is raised, would need to be enforced.”
The next step in development will be to ‘teach’ the new system about the eruption history at New Zealand’s other
volcanoes including Tongariro and Ruapehu which both attract tens of thousands of visitors a year.
The events at Whakaari are the subject of a number of reviews mandated under legislation and involve a number of
Government agencies and Departments.
The research and development of the new forecasting system is published in Nature Communications.