Professor Cunwei Lu of the Faculty of Engineering, Fukuoka Institute of Technology has begun a demonstration experiment of tsunami monitoring using three-dimensional (3D) image analysis and AI (artificial intelligence) analysis technology in the Genkai Nada off the coast of Fukuoka Prefecture.A high-performance camera accurately captures an offshore tsunami, and AI analyzes 3D images. If it is put into practical use, accurate tsunami prediction will be possible 3 to 20 minutes before the arrival of the tsunami.
According to Fukuoka Institute of Technology, in the demonstration experiment, two cameras installed on the roof of the Fukuoka Institute of Technology research building use infrared and visible light cameras together to monitor the Genkai Nada during the day and at night. AI compares the sea level and wave data in normal times, analyzes the characteristics of wave height and shape, and automatically determines whether it is a general wave, typhoon storm surge, or tsunami.
In the demonstration experiment, a system is in place to predict a tsunami 20 km away during the day and 10 km away at night.When the observation data obtained in the study was compared with the actual data held by the Japan Meteorological Agency, the average error was only about 20 cm.Professor Roh believes that by combining the construction of a camera tsunami observation network and wave height analysis using a microwave radar, it is possible to predict a tsunami 50 kilometers away.
The tsunami is thought to cause great damage even in the Nankai Trough earthquake, which is expected to occur with high probability, including swallowing 2011% of the deaths of more than 1 people in the 5,000 Great East Japan Earthquake.
However, no technology has been established to accurately predict when a tsunami will occur after an earthquake and on what scale it will hit the land.
reference:[Fukuoka Institute of Technology] Artificial intelligence discovers a tsunami 50 kilometers away
