Sign In

Communications of the ACM

ACM TechNews

AI Can Spot Early Signs of a Tsunami from Atmospheric Shock Waves

View as: Print Mobile App Share:
Flooding caused by a tsunami that hit the coast of Sri Lanka in 2004.

A truly effective tsunami detection system would require international cooperation to share data from satellite constellations.

Credit: DigitalGlobe/Getty Images

Researchers at Florida-based satellite manufacturing company Terran Orbital Corp. found that off-the-shelf artificial intelligence (AI) models can detect the early signs of a tsunami in two-dimensional (2D) images from GPS satellites.

The researchers used data generated by a computer algorithm developed by researchers at NASA's Jet Propulsion Laboratory and Italy's Sapienza University of Rome, which measures changes in the density of charged particles in the ionosphere as tsunamis form.

The data was transformed into 2D images that were analyzed by the AI to identify features associated with tsunamis.

The AI achieved a reported detection performance rate above 90% after eliminating ionospheric disturbance patterns that at least 70% of ground stations in contact with the satellites failed to pick up.

From New Scientist
View Full Article - May Require Paid Subscription


Abstracts Copyright © 2023 SmithBucklin, Washington, D.C., USA


No entries found