The Tech Behind Tsunami Alerts: How AI and Sensors Save Lives

If you live in a landlocked city, tsunamis may seem like something out of a movie. But for the hundreds of millions of people living in areas vulnerable to ocean hazards— like Alaska, Chile, Japan or Indonesia—tsunamis are a very real threat. Most recently, on July 30th, a magnitude 8.8 earthquake off Russia’s Kamchatka Peninsula triggered a small tsunami. But we’ve seen devastating events in recent history, like the 2011 tsunami in Japan and the 2004 Indian Ocean tsunami, which claimed over 230,000 lives across multiple countries. The 2004 disaster, which struck with no warning system in place, became a global wake-up call—underscoring the urgent need for early detection and coordinated response systems. Today, tsunami centres around the world rely on technology to monitor earthquake activity and warn when needed. So, what exactly happens when a tsunami threat is detected—and what role does technology play in the process? Technology plays a fundamental role in warning. Without it, authorities would be blind. In 80% of tsunami cases, it all starts with an earthquake. And as early as that, technology is at play. As an earthquake happens, seismic sensors or stations, both in the ocean and on the ground, pick up the activity. They sense the movement of the ground, acceleration or even acoustic signals. There’s also seafloor pressure sensors and surface buoys, which measure sea level to estimate threats. Sensors, powered by batteries, solar, wind, or all of the above, send real-time data to a monitoring centre through radio, satellite or cellular connectivity and there, the data is analyzed. The data is also checked to find the epicenter and magnitude. Simulations are made to estimate the height waves could reach, the areas affected and when they could arrive to see if there’s a threat of tsunami. Basically, the hardware, so sensors, sends the data to the software, in the tsunami center, where calculations are made. There is a lot of hardware out there now. The Pacific Tsunami Warning Center for example has some 600 seismic stations around the world to spot tremors that could trigger tsunamis. Japan has a very sophisticated system in which ground sensors, as well as underwater ones, gather data on earthquakes and possible changes in the water. These are part of an international effort, as agencies, countries and regions have their own networks and systems and share the data. Now, to the calculations. Most tsunami warning centers today still rely on deterministic, physics-based simulations like NOAA’s MOST or Japan’s TUNAMI-N2. But now, AI is starting to show how it could help with simulations for more accurate predictions of possible tsunamis and their effects. For example, the Tohoku and Tokyo Universities and Fujitsu developed an AI model with the Fugaku supercomputer in 2021 to predict tsunami flooding in near real time. Researchers at UCLA and Cardiff University are training AI acoustic classifiers to identify tsunami-generating events from hydrophone signals. And researchers at the University of Texas at Austin recently presented a digital twin that uses acoustic pressure data from seafloor sensors to forecast tsunami propagation. Thanks to all this technology and after the threat is confirmed through the data and models, emergency agencies, are notified and a message is put out for the public. The possible tsunami is monitored and updated through real-time data from buoys, cameras or tide gauges. The key to reducing the tsunami’s impact lies in early detection and fast, accurate communication. It’s a collaborative effort, in which data is constantly shared between different tsunami agencies. Tsunamis can’t be prevented, but, their effects can be reduced, especially when it comes to human life loss when there’s early warning. Technology like sensors, AI or digital twins, are playing a key role in giving countries a chance to not be caught off guard and avoid repeating history.