Scientists at Lawrence Livermore National Laboratory (LLNL) have contributed to the development of a real-time tsunami forecasting system, utilizing El Capitan, recognized as the world’s fastest supercomputer. The new system aims to enhance early warning capabilities for coastal communities near earthquake zones.
El Capitan, developed with support from the Advanced Simulation and Computing (ASC) program at the National Nuclear Security Administration (NNSA), has a theoretical peak performance of 2.79 quintillion calculations per second. Researchers used this computing power in an offline precomputation phase before El Capitan transitions to classified national-security work. The effort generated a large library of physics-based simulations that connect earthquake-induced seafloor motion to resulting tsunami waves.
The project relied on more than 43,500 AMD Instinct MI300A Accelerated Processing Units (APUs) to address complex acoustic-gravity wave propagation problems. This process produced data enabling real-time tsunami forecasting on smaller systems by performing computationally intensive tasks in advance. According to researchers, this approach allows for rapid predictions during actual tsunamis using modest GPU clusters.
The system was developed in collaboration with the Oden Institute at the University of Texas at Austin and the Scripps Institution of Oceanography at the University of California, San Diego. It uses “digital twin” models that simulate seafloor earthquake motion based on real-time pressure sensor data and advanced physics-based simulations.
“This is the first digital twin with this level of complexity that runs in real time,” said LLNL computational mathematician Tzanio Kolev, co-author on the paper. “It combines extreme-scale forward simulation with advanced statistical methods to extract physics-based predictions from sensor data at unprecedented speed.”
By using El Capitan for offline precomputation, researchers solved a billion-parameter Bayesian inverse problem in less than 0.2 seconds, achieving prediction speeds much faster than existing methods.
Researchers believe this capability could significantly improve emergency response and potentially save lives by forming part of next-generation early warning systems. For example, during 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 tsunami warning systems often use seismic and geodetic data but rely on simplified models that may not capture fault rupture complexities, sometimes leading to false alarms or late warnings. The new approach instead uses seafloor pressure sensors and solves full-physics models rapidly.
As sensor networks expand along earthquake-prone coasts and computational resources improve, researchers anticipate broader deployment of this method in future warning systems.
“This framework represents a paradigm shift in how we think about early warning systems,” said senior author Omar Ghattas, professor of mechanical engineering and principal faculty in the Oden Institute at UT-Austin. “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.”
The MFEM open-source finite element library developed by LLNL enabled scalable GPU-accelerated simulations for this project. Using 43,520 APUs on El Capitan, 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 the Department of Energy’s (DOE) 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 explained that after precomputations are complete, forecasting tsunami wave heights can be performed on smaller GPU clusters due to efficient algorithms designed 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,” said Kolev. “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.”
Kolev added that their Bayesian inversion framework could be applied beyond tsunamis—to wildfire tracking, subsurface contaminant monitoring, space weather forecasting and intelligence applications requiring quick decisions based on data analysis.
Project collaborators 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.
