February 21, 2024

Drawing on arms control technology and artificial intelligence, a team of scientists from Cardiff University’s School of Mathematics has developed a method that uses underwater microphones to provide early warning of potentially deadly tsunamis.

Tsunamis, the result of earthquakes, volcanic eruptions, underwater landslides and other causes, are among the deadliest natural disasters, capable of causing widespread destruction tens of thousands of miles away. In 2004, a storm hit the Indian Ocean, killing an estimated 230,000 people as massive walls of displaced water struck Indonesia, Sri Lanka, India, Thailand, Somalia, Myanmar, Maldives, Malaysia, Tanzania, Seychelles, Bangladesh, South Africa, Yemen , and Kenya.

There are tsunami detection and warning systems, although these systems leave a lot to be desired. Based on seismometers and bottom pressure sensors attached to the buoys, they can detect earthquakes, but not all earthquakes cause tsunamis, and the buoys can only detect tsunamis as they pass by, leaving little time to react.

These warning systems are useful, but very limited. The formation and propagation of tsunamis is due to a very complex interaction of many factors. That’s why some can destroy entire areas, while others may only raise the water level a few feet when they come ashore.

To improve on this, the Cardiff team developed a mathematical model based on data gathered from the network of hydrophones established to enforce the 1996 Comprehensive Nuclear-Test-Ban Treaty, by cocking electronic ears to pick up the unique sea of ​​nuclear blasts sound. While listening, the underwater microphones were able to detect four tsunami-related earthquakes.

Using this data, advanced acoustic techniques and artificial intelligence, the team was able to detect and analyze the sound radiated by 200 earthquakes in the Pacific and Indian Oceans in real time. They were able to locate the origin of the earthquake, describe the resulting pressure field, the duration of the wave and the speed at which it traveled. This allows the system to classify the type and magnitude of the earthquake and the size of the tsunami.

This information not only saves lives, but also helps avoid false positives and tailor warnings based on predicted danger. Designed to work with existing warning systems, the next step will be to develop user-friendly software that could be installed at national warning centers sometime this year.

“Our research shows how reliable information on the size and magnitude of a tsunami can be quickly obtained by monitoring acoustic gravity waves, which travel much faster through water than tsunami waves, giving people more time to evacuate places before landfall ,” said Dr. Osama Kadri, Senior Lecturer in Applied Mathematics.

The study was published in fluid physics.

source: Cardiff University