UCSD seismologist wins supercomputing prize for tsunami prediction work

Seismologist Alice Gabriel of UC San Diego in La Jolla is part of a multi-institution team that won supercomputing’s most prestigious prize Nov. 20 for its studies on a real-time tsunami monitoring prototype that could greatly improve early-warning systems.

Gabriel, who works in UCSD’s Scripps Institution of Oceanography, and her colleagues from the University of Texas at Austin and Lawrence Livermore National Laboratory won the 2025 Association for Computing Machinery Gordon Bell Prize, often referred to as the Nobel Prize of supercomputing.

The award, given at the annual Supercomputing Conference, recognizes “innovation in high-performance computing applications in science, engineering and large-scale data analytics,” according to the prize’s website.

The team also received the Hyperion Research HPC Innovation Excellence Award and the HPCWire Reader’s Choice Award for Best Use of HPC in Physical Sciences.

The research used some of the world’s largest supercomputers in a “digital twin” model to simulate the seismically active Cascadia subduction zone of the Pacific Northwest in such a complex manner that it reduces the time needed to calculate the variables from roughly 50 years to less than a second.

A digital twin is a computer replica of the real geophysical system that combines physics models with real-world observations to estimate the changing variables and explore plausible futures.

“Using the largest supercomputers in the world, we developed new algorithms using seafloor sensors to forecast tsunami wave heights and their uncertainties in fractions of a second — 10 billion times faster than conventional algorithms,” said Gabriel, who also was nominated for the Bell Prize in 2014. “Our work shows that physics-based models of earthquakes and tsunami generation are now fast enough to guide real-time response, which not only improves early warning for the Pacific Northwest but also points the way toward global, physics-based earthquake and tsunami early-warning systems.”

Such a system, which could be put in place with as few as 600 pressure sensors along the length of the subduction zone — which spans 700 miles between Northern California and British Columbia — would dramatically reduce tsunami warning times by monitoring changes on the ocean floor from any seismic event.

Gabriel said it can take 10-15 minutes for scientists to recognize a probable tsunami event and disseminate that information to the public, like what happened in July when an 8.8 magnitude earthquake rocked the Kamchatka Peninsula in Russia and tsunami notices were issued as far away as San Diego County.

For the Pacific Northwest fault, just 100 miles off the coast, reducing those 10-15 minutes could have lifesaving consequences. The most recent major Cascadia earthquake was in 1700, and seismologists have estimated a 37% probability of a magnitude 8.0 or higher earthquake in the next 50 years.

“It is overdue,” said Gabriel, who has been at Scripps Oceanography since 2022. “The science community has been pushing for a more robust monitoring system for almost a decade.”

Expense is the most prohibitive factor, she said, but before her team’s work, the methods to measure such movements were relatively untested.

Applications from the research could enable early-warning systems in countries with lengthy coastlines and fewer monitoring stations than the United States, as well as uses in other fields such as wildfires, seismic ground motion and storm surges.