New GPS-based method flagged the exact patch of fault that later ruptured in Kamchatka's magnitude 8.8 quake

Geophysicists at the University of California, Riverside have developed an algorithm that uses GPS measurements of slow ground deformation to identify which sections of a fault are locked and accumulating the strain that eventually powers the largest earthquakes on Earth. Applied retroactively to the Kamchatka subduction zone in far eastern Russia, the method highlighted a stretch of the fault that later ruptured in the magnitude 8.8 earthquake of July 29, 2025, one of the most powerful earthquakes ever recorded by modern instruments.

The study, led by UC Riverside geophysicists Gareth Funning and Axel Periollat, was published in the American Geophysical Union journal Geophysical Research Letters and described in a university announcement on September 4. The paper, titled "Linking Interseismic Locking to Coseismic Rupture: The 2025 Mw 8.8 Kamchatka Earthquake," does not claim to forecast when an earthquake will happen. Instead, it argues that the pattern of ground motion recorded by GPS stations in the years before a rupture can reveal where a fault is locked, and therefore where the next large rupture is most likely to begin and spread.

Reading strain in slow ground motion

Subduction zones, where one tectonic plate slides beneath another, produce the planet's largest earthquakes because the two plates can lock together across enormous patches of the fault interface rather than sliding smoothly past one another. Those locked patches, known to seismologists as asperities, behave like zones of friction that resist motion while the surrounding plate boundary keeps converging, storing elastic energy for years or decades until the patch finally slips and releases it all at once.

Funning and Periollat's approach uses networks of continuously operating GPS receivers to track the almost imperceptible surface deformation caused by that locking, then works backward to map which parts of the underlying fault plane are coupled tightly enough to be storing significant strain. "Earthquakes capture headlines when they happen, but for years beforehand the fault is quietly accumulating strain. This strain can be measured," Funning said, according to the university's account of the research.

Kamchatka provided an unusually direct test. The peninsula sits along one of the most active stretches of the Pacific "Ring of Fire," where the Pacific plate dives beneath the Okhotsk plate, and it has a long instrumental and historical earthquake record, including a magnitude 9.0 event in 1952. When the researchers ran their locking analysis using GPS data collected before the 2025 earthquake, the model flagged a patch of the subduction interface beneath the peninsula as strongly coupled. Months later, the actual rupture broke through close to that same patch, an event the U.S. Geological Survey's earthquake catalog lists as tying for the sixth-largest ever recorded by modern instruments. "We had an idea where the strain was accumulating based on a relatively limited data set," Periollat said. "Seeing it work so well confirmed that this approach has real potential."

What the method can and cannot tell scientists

The July 2025 Kamchatka earthquake triggered tsunami warnings across the Pacific, though the wave that ultimately reached shorelines proved smaller than the one generated by the 1952 earthquake in the same region. Funning and Periollat's paper attributes that difference partly to how much the shallowest part of the fault, closest to the seafloor and most efficient at displacing water, actually slipped, a detail that falls outside what a locking map alone can predict.

The researchers are explicit that their technique addresses location, not timing or severity. It cannot say how large a future earthquake will be, whether it will generate a damaging tsunami, or when within a period of years or decades it might occur. "Your peace of mind shouldn't come from believing we can forecast the exact earthquake," Funning said. What the method offers instead is a way to prioritize which segments of a hazardous fault deserve the closest monitoring and the most conservative building codes, narrowing an otherwise enormous search area down to the patches doing the most work.

That distinction matters because seismology has a long history of failed short-term prediction schemes, and researchers in the field are generally careful to separate probabilistic hazard assessment, which describes long-term likelihoods for a region, from precise forecasting of a specific event. Interseismic coupling models, built from decades of geodetic data across subduction zones from Chile to Japan, have already shaped hazard maps used by engineers and emergency planners, and independent coverage of the new study notes that the Kamchatka case adds a specific, retrospectively verified example of a coupling map aligning with where a real megathrust rupture subsequently occurred.

Extending the approach to other fault systems

UC Riverside researchers say they are now applying the same locking analysis to other subduction zones with dense GPS coverage, including sections of the plate boundaries beneath Japan, Mexico, New Zealand and the Pacific Northwest, where the Cascadia subduction zone is capable of producing its own magnitude 9 earthquake. The team is also examining California's Hayward Fault, an inland strike-slip fault running through the East Bay that has a mix of creeping sections, which slide continuously and release strain gradually, and locked sections that could rupture abruptly.

A persistent limitation is the scarcity of measurements offshore, where much of a subduction zone's locked interface actually lies beneath the seafloor rather than under land-based GPS stations. Japan has installed networks of seafloor acoustic-geodetic instruments to fill that gap, and researchers elsewhere, including in Chile and the Pacific Northwest, have proposed similar offshore arrays. Newer satellite missions capable of measuring ground deformation from orbit may eventually supplement sparse ground networks in remote regions such as Kamchatka, giving the coupling-mapping approach a fuller picture of where strain is building along the world's most hazardous plate boundaries.