Scientists at Lawrence Livermore National Laboratory (LLNL) have developed a new real-time tsunami forecasting system using El Capitan, currently the world’s fastest supercomputer. The initiative aims to improve early warning systems for coastal communities near earthquake zones.
El Capitan, funded by the Advanced Simulation and Computing program at the National Nuclear Security Administration, can reach a theoretical peak performance of 2.79 quintillion calculations per second. Before its transition to classified national-security work, LLNL researchers used El Capitan in an offline precomputation phase to generate a large library of physics-based simulations connecting earthquake-induced seafloor movement to resulting tsunami waves.
The project involved more than 43,500 AMD Instinct MI300A Accelerated Processing Units (APUs) to address large-scale acoustic-gravity wave propagation problems. This process produced data that enables real-time tsunami forecasting on smaller computing systems. By conducting intensive computations upfront on El Capitan, the team solved a high-fidelity Bayesian inverse problem that allows rapid predictions during actual tsunamis using modest GPU clusters.
Development partners included the Oden Institute at the University of Texas at Austin and the Scripps Institution of Oceanography at the University of California, San Diego. The resulting digital twin models use real-time pressure sensor data and advanced simulations to infer earthquake impacts on the ocean floor and forecast tsunami behavior with uncertainty quantification.
“This is the first digital twin with this level of complexity that runs in real time,” said Tzanio Kolev, computational mathematician at LLNL and co-author of the paper. “It combines extreme-scale forward simulation with advanced statistical methods to extract physics-based predictions from sensor data at unprecedented speed.”
Using El Capitan’s computing power for precomputation enabled researchers to solve a billion-parameter Bayesian inverse problem in less than 0.2 seconds, predicting tsunami wave heights much faster than previous methods.
Researchers noted that this capability could significantly improve emergency response efforts and save lives by providing next-generation early warning systems. For events such as a magnitude 8.0 or larger earthquake along the Cascadia Subduction Zone in the Pacific Northwest, destructive waves could reach shore within ten minutes—leaving little time for evacuation.
Current warning systems often depend on seismic and geodetic data but may use simplified models that do not capture fault rupture complexities, leading to false alarms or late warnings. The new approach uses seafloor pressure sensors and solves full-physics models quickly.
As sensor networks expand along vulnerable coastlines and computational resources advance, researchers see potential for deploying this method in future warning systems for faster and more reliable alerts.
“This framework represents a paradigm shift in how we think about early warning systems,” said Omar Ghattas, professor of mechanical engineering at UT-Austin and senior author of the study. “For the first time, we can combine real-time sensor data with full-physics modeling and uncertainty quantification — fast enough to make decisions before a tsunami reaches the shore. It opens the door to truly predictive, physics-informed emergency response systems across a range of natural hazards.”
Central to this system is MFEM, LLNL’s open-source finite element library enabling scalable GPU-accelerated simulations including acoustic-gravity wave propagation in oceans. On El Capitan’s 43,520 APUs, MFEM handled simulations involving 55.5 trillion degrees of freedom—a record for unstructured mesh finite element simulation.
“MFEM’s high-order methods and GPU readiness, developed under the ASC program at LLNL and Department of Energy’s Exascale Computing Project, made it possible to scale to the full machine,” Kolev said. “This was really a first-of-its-kind demonstration of how we can use that power not just for raw performance, but also for mission-relevant, time-critical decisions in many MFEM-based applications.”
Kolev added that after completing precomputations on El Capitan, online steps such as inferring seafloor motion and forecasting tsunami wave heights can be done on smaller GPU clusters due to algorithm design optimized for GPUs.
“This work is important because it shows that we can solve an inverse problem of enormous size — not for 10 or 15 variables, but for millions, or even billions of variables very quickly,” Kolev said. “In the past you’d either have a fast model that’s not accurate or a full-physics model that takes hours or days. Now we’re showing that we can do both — accurate and fast — using principled mathematics and modern computing.”
He also stated that this Bayesian inversion framework could be adapted beyond tsunamis for other complex scenarios such as wildfire tracking or space weather forecasting where rapid data-driven decisions are necessary.
The research team included Veselin Dobrev and John Camier from LLNL; Omar Ghattas, Stefan Henneking, Milinda Fernando and Sreeram Venkat from UT-Austin; and Alice-Agnes Gabriel from UC San Diego.
