AI Used For Developing Tsunami Early Warning System To Classify Earthquakes, Determine Risk

Researchers have tried using AI to develop a tsunami early warning system that can classify earthquakes and evaluate potential tsunami risks.

Devastating Tsunamis

The dangers and damage that tsunamis bring are hardly debatable. These natural disasters are among the world's most fearsome events.

These events are so destructive that there was even one tsunami that led to the deaths of over 200,000 people. According to USA Today, this took place in Asia in 2004. Across the US, strong tsunamis have also occurred and are likely to happen again.

While tsunamis bring killer and extremely destructive waves, they have been incredibly hard to predict. This is due to their risk being strongly dependent on the underwater earthquake features, as noted by Science Daily. This also means that it is difficult to issue early warnings for such devastating events.

Tsunami Early Warning System

Now, researchers from the University of California and Cardiff University have been able to collaborate and come up with a tsunami early warning system. Their system mixed artificial intelligence with acoustic technology to facilitate immediate earthquake classification and tsunami risk evaluation. Their paper was published in the Physics of Fluids publication.

As per Science Daily, underwater earthquakes may trigger tsunamis in cases of water displacement in large quantities. Hence, being able to classify earthquakes is vital to gauging tsunami risk.

Bernabe Gomez, one of the study co-authors, explains that tectonic events that have a strong element of vertical slips likely elevate or lower the water column. This is in comparison to elements that are horizontal-slip. Hence, being able to identify the type of slips early on can help decrease false alarms. This may also boost warning reliability.

In such instances, time is a primary concern. Science Daily adds that depending on wave buoys situated in the deep ocean often leads to insufficient time for evacuation.

Hence, the researchers proposed to gauge acoustic radiation that earthquakes produce. This phenomenon holds insights regarding tectonic occurrences. Not to mention, they travel much faster compared to strong tsunami waves. Acoustic radiation may be recorded, gauged, and monitored via hydrophones, or underwater microphones, in real-time.

Usama Kadri, another co-author of the study, explains that the travel time of acoustic radiation is less compared to that of tsunami waves. It also holds vital data regarding its origins. Moreover, its field of pressure can also be picked up even if it were a thousand kilometers away from its source. Kadri notes that being able to derive analytical solutions for such fields of pressure is vital when it comes to analysis in real-time.

While the earthquake source is triangulated from the hydrophones by the computational model, algorithms of AI help in identifying its magnitude and slip type. Other vital properties, such as effective width and length, duration, and uplift speed, also get calculated.

The researchers tried out their model using hydroponic data that is available. By doing so, they observed how the system was able to immediately and successfully describe the parameters of the earthquake using minimal computational requirements.

These scientists are now working towards enhancing their model by taking other factors into account. This may, in turn, boost the accuracy of characterization.

Such endeavors are part of overarching endeavors to boost warning systems for hazards.

Check out more news and information on Natural Disasters in Science Times.