New volcano alert system can predict eruptions 16 hours ahead, give people time to evacuate and help save lives
The new alert system could have averted the 2019 disaster in New Zealand that killed 21 tourists, researchers said
New Zealand scientists have created a volcano warning system to predict a future disaster. The step comes after the country witnessed a horrific event last year when an eruption in the active volcano island called Whakaari or the White Island killed 21 tourists.
Over the last decade, volcanic eruptions and the resulting hazards, such as tsunami, have claimed more than 1,000 lives. New Zealand does not have an advanced warning system. Instead, the country depends on a volcano alert level (VAL) system, which gathers information on the state of the volcano. Researchers analyze this data and make predictions after reaching a consensus. Besides, the system updates information every few weeks or months and does not provide real-time information.
The new alert system is automated and can predict an eruption 16 hours ahead, giving people time to evacuate. It could have averted the 2019 disaster, the researchers said. "Our automated warning system provides real-time hazard information and a much greater level of safety to protect tourists and help operators determine when it is safe to visit volcanoes," the researchers wrote in The Conversation.
“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 in future, but it’s a question of whether policymakers, landowners and tourist operators want a system that, once an alert is raised, would need to be enforced," Professor Shane Cronin from the University of Auckland, explained in a statement.
The new warning system is based on machine-learning. Researches fed the system with ten years of volcano data recorded from the White Island. They then trained it to identify patterns associated with eruptions. Based on the data, it predicted four out of five events. It missed detecting the fifth one because the data pattern was different. “Machine learning means it learns from ‘experience’ and so constantly improves its accuracy,” Dr. Andreas Kempa-Liehr from the Faculty of Engineering, explained.
The system warns researchers of an eruption when there is a one in 12 chances of it occurring. The alert stays for five days, which means the island could be out-of-bounds for a month every year. “This system detects eruption types that are most likely to be fatal,” Professor Cronin noted.
The system has been up and running for five months now. Currently, the researchers are "working on how best to integrate this to strengthen their existing protocols and provide more timely warnings at New Zealand volcanoes," they wrote in The Conversation.
So far, the system has been trained using data from White Island. They are now hoping to feed data on other volcanoes such as Tongariro and Ruapehu, both of which are popular tourist destinations.
But the system might not be able to forecast volcanic eruptions if the pattern is different from the usual ones. "In any case, human operators, whether assisted or not by early warning systems, will continue to play an important role in safeguarding those living near or visiting volcanoes," the researches added.
The study is published in Nature Communications.