[ad_1]
The killer waves came with little warning.
One of the deadliest natural disasters in recorded human history – a tsunami that killed more than 200,000 people in Asia in 2004 – took most victims by complete surprise. Closer to home, large tsunamis have occurred in the U.S. and will undoubtedly occur again, the U.S. Geological Survey warned.
Worldwide, tsunamis remain among nature’s most fearsome natural disasters.
Now, thanks in part to artificial intelligence, scientists are working on a new early warning system to provide coastal residents with more lead time ahead of a potentially deadly tsunami.
According to a new study published Tuesday in the journal Physics of Fluids, an early warning system is being developed that combines “state-of-the-art acoustic technology with AI to immediately classify earthquakes and determine potential tsunami risk,” the study said.
AI in the news:AI-generated images already fool people. Why experts say they’ll only get harder to detect.
‘Highly destructive events’
“Tsunamis can be highly destructive events causing huge loss of life and devastating coastal areas, resulting in significant social and economic impacts as whole infrastructures are wiped out,” said study co-author Usama Kadri of Cardiff University in Wales.
The new research used sound recordings captured by underwater microphones, called “hydrophones,” to measure the acoustic radiation produced by 200 earthquakes that happened in the Pacific and Indian Oceans.
“Our study demonstrates how to obtain fast and reliable information about the size and scale of tsunamis by monitoring acoustic-gravity waves, which travel through the water much faster than tsunami waves, enabling more time for evacuation of locations before landfall,” Kadri said.
What if the concierge is a robot?:How artificial intelligence will change your hotel stay.
What type of earthquake is it?
Underwater earthquakes can trigger tsunamis if a large amount of water is displaced, so determining the type of earthquake is critical to assessing the tsunami risk.
There are two types of underwater earthquakes, and only one can produce tsunamis, researchers say.
“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 of the University of California, Los Angeles. “Thus, knowing the slip type at the early stages of the assessment can reduce false alarms and enhance the reliability of the warning systems,” Gomez added.
Kadri told USA TODAY that “our machine-learning model was trained with 200 real underwater earthquakes, and is capable of classifying the slip mode: horizontal (which does not produce tsunamis), or vertical movement (which results in tsunamis.)”
Will AI replace your doctor?:As a physician, I’m worried new tech will hurt patient care.
Warning systems can result in false alarms
Kadri said warning systems rely on decisions that are mainly based on the earthquake magnitude and location, which could result in false alarms. “In that sense, the current warning systems are rather conservative and not so much reliable,” Kadri said.
The new machine-learning model is able to analyze the hydrophone (underwater microphone) data “within a few seconds on a standard computer.”
In addition, systems that rely on deep ocean wave buoys to measure water levels often leave insufficient evacuation time.
Designed to be used alongside existing warning systems, the new system triangulates the source of the earthquake from the hydrophones and AI algorithms classify its slip type and magnitude. It then calculates important properties like effective length and width, uplift speed, and duration, which dictate the size of the tsunami.
New system set to be used worldwide
This new research predicting tsunami risk is part of a long-running project to enhance natural hazard warning systems across the globe, according to a statement from Cardiff University.
Their latest development is featured in user-friendly software that is set to be hosted in national tsunami warning centers later this year.
[ad_2]
Source link
