The 2024 Indian Ocean tsunami caused mega-damage in Aceh, and Thailand and claimed more than 200,000 lives. Although the earthquake was detected early, the movement and strength of the tsunami generated could not be predicted by scientists at the time. More accurate and faster predictions are needed to prevent this tragedy from happening again.
Researchers from Lawrence Livermore National Laboratory (LLNL) have now announced the success of developing a real-time tsunami prediction system that is not only more accurate but also does it very quickly.
The power of the world's fastest supercomputer, El Capitan, is used to perform inverse Bayesian approximations with more than 1 billion parameters in just 0.2 seconds. Because of El Capitan's high power, tsunami wave predictions are 10 billion times faster than previous capabilities.
Data from seafloor pressure sensors collected in real time will be processed by El Capitan quickly, allowing areas predicted to be hit by tsunamis to be evacuated earlier. In other words, El Capitan allows a digital twin of the tsunami to be created which can then be tracked by authorities before emergency response plans are activated.
A preprint of the research paper has been published in the journal arXiv