Tsunami Prediction, Explained

Tsunami early warning systems today rely on models that can take precious minutes to run - minutes that coastal communities don't have. New research is changing that by building a "digital twin" of the earth's crust, one that can simulate the physics of an earthquake and its resulting tsunami in near-real time using some of the most powerful supercomputers on the planet.

In this talk, Tzanio Kolev will walk through how his team used Bayesian inversion techniques on the El Capitan supercomputer to model the Cascadia subduction zone - the fault line that threatens the Pacific Northwest with a potentially catastrophic earthquake and tsunami. This work, recognized with the 2025 ACM Gordon Bell Prize, represents a major leap in how we use extreme-scale computing to protect lives.

Join us for a look inside the intersection of high-performance computing, applied mathematics, and disaster preparedness - and what it takes to simulate the planet fast enough to matter.