Scientists have developed an artificial intelligence-powered early warning system for tsunamis and underwater earthquakes.
A team at Cardiff University has been able to monitor the movement of tectonic plates in real time, combining technology such as underwater microphones with artificial intelligence.
Experts have used the recordings to detect 200 earthquakes in the Pacific and Indian Oceans.
Study co-author Dr Usama Kadri, Senior Lecturer in Applied Mathematics at Cardiff University, said: “Our study shows how monitoring acoustic gravity waves can quickly yield reliable information about the size of a tsunami.
“[They] Traveling through water is much faster than tsunami waves, giving people more time to evacuate before making landfall.”
Acoustogravity waves are sound waves that travel through the deep ocean at the speed of sound. They can be created by underwater earthquakes, explosions and landslides – all of which can lead to tsunamis.
The new system uses underwater microphones or hydrophones, recordings and “computational models” to triangulate the source of tectonic events, even thousands of kilometers away.
Mr Kadi and his partner Dr Bernabe Gomez Perez currently have 11 hydrophones in oceans around the world. With 24, he told us, they could monitor the entire planet.
read more:
German publisher apologizes to Michael Schumacher’s family over AI-generated interview
Google boss admits AI dangers ‘keep him up at night’
Existing warning systems rely on waves reaching buoys and seismic sensors at sea to trigger tsunami warnings.
This leaves little time for evacuation, where a few extra minutes could be the difference between life and death.
However, they are not always accurate in predicting the danger posed by tsunamis.
AI systems combined with acoustic technology can detect epicenters of tectonic movement and predict where evacuations will be required.
Mr Quadri told Sky News: “Current systems can’t calculate in real time and don’t have any real tools to measure tsunamis – and because the buoys are so close to shore, it’s often too late.
“Take the example of the tsunami in Sri Lanka [Indian Ocean] In 2004, suppose we could give people 65 minutes to evacuate. This is the possibility to save almost all lives. “
“Once the wave hits the hydrophone, the calculation takes about 17 seconds. Even we were surprised by the speed.”
He explained that their system is designed to work with existing systems, with the two balancing each other.
The new system’s algorithms can classify an earthquake’s “slip type” and magnitude before analyzing its properties, such as length and width, uplift rate and duration.
This can be used to find out the actual size of the tsunami.
Co-author Dr Gomez Perez, now at the University of California, Los Angeles, said: “Tectonic events with strong vertical sliding elements are more likely to raise or lower the water column than horizontal sliding elements.
It can also predict how long it will take for a tsunami to reach a selected location (past example).
“Thus, knowing the type of slippage at an early stage of assessment can reduce false positives and complement and enhance the reliability of early warning systems through independent cross-validation.”
Mr Cardy said any false positives could shut businesses and ports and have significant financial repercussions.
“If you’re crying wolf, people will lose faith in the system. Then when it’s real, people won’t leave,” he added.
The team has been meeting with the Oceanographic Commission of UNESCO to discuss opportunities to use the technology for disaster preparedness. Portugal, known for its giant waves, was particularly interested in the insights it could provide.
The team’s work to predict tsunami risk is part of a long-term project to strengthen early warning systems for natural hazards around the world.
The full report appears in the April 25, 2023 issue of Physics of Fluids.