Tsunamis can be incredibly destructive, causing huge loss of life and destroying infrastructure. In the past 100 years, 58 tsunamis have claimed more than 260,000 lives, or an average of 4,600 per disaster, more than any other natural hazard, according to data from the United Nations in November 2022.
Early warnings for tsunamis are crucial to preventing such losses. A new study has developed an early warning system using artificial intelligence (AI) to classify earthquakes and determine tsunami risk.
Early warnings for tsunamis are difficult because the risk is highly dependent on the features of the underwater earthquake that triggers it. These earthquakes can cause tsunamis if a large amount of water is displaced. Hence, determining the type of earthquake is important.
Researchers from the University of California, Los Angeles and Cardiff University in the UK have developed an early warning system that combines state-of-the-art acoustic technology with AI to immediately classify earthquakes which can help identify potential tsunami risk, according to a press statement published by AIP Publishing.
“Tectonic events with a strong vertical slip element are more likely to raise or lower the water column compared to horizontal slip elements,” said co-author Bernabe Gomez. “Thus, knowing the slip type at the early stages of the assessment can reduce false alarms and enhance the reliability of the warning systems through independent cross-validation.”
In this study, published in the journal Physics of Fluids, the researchers focused on measuring the acoustic radiation (sound) produced by the earthquake, which carries information about the tectonic event and travels significantly faster than tsunami waves, according to the statement. Hydrophones or underwater microphones recorded the acoustic waves and monitor tectonic activity in real-time.
Acoustic radiation travels through the water column significantly faster than tsunami waves. Furthermore, it carries information about the origin and its pressure field can be recorded at distant locations, said co-author Usama Kadri in the statement. “The derivation of analytical solutions for the pressure field is a key factor in the real-time analysis,” Kadri added.
The computational model triangulates the source of the earthquake from the hydrophones and AI algorithms classify its slip type and magnitude, according to the statement. Crucial properties such as effective length and width, uplift speed, and duration, which dictate the size of the tsunami, are calculated.
The researchers used available hydrophone data for testing and found it almost instantaneously and successfully described the earthquake parameters with low computational demand. The researchers are further improving the model adding more information to increase the tsunami characterization’s accuracy. These findings are part of a larger project to improve hazard warning systems.
