The San Andreas Fault runs roughly 800 miles through California, marking the boundary where the Pacific and North American tectonic plates grind past one another. Most of the fault creeps steadily or ruptures on a semi-regular cycle, but one stretch near the Salton Sea in Southern California has remained locked since sometime around the year 1690, quietly storing more than three centuries of accumulated strain with no major release. Seismologists consider this segment one of the most closely watched fault lines in the world, precisely because of how overdue it appears relative to its own history.
Two Plates Grinding Past Each Other for 800 Miles
The San Andreas is a transform fault spanning roughly 1,300 kilometers, where the Pacific plate and the North American plate slide horizontally past each other rather than colliding or pulling apart. The fault behaves very differently along its length. Near the town of Parkfield in central California, a short creeping section releases moderate, roughly magnitude-6 earthquakes with unusual regularity, six such quakes have struck the area since 1857, spaced about 22 years apart on average, according to USGS research on the segment. On either side of that creeping stretch, the fault locks for centuries at a time: the section north of San Juan Bautista ruptured in the magnitude-7.9 San Francisco earthquake of 1906, an event that killed an estimated 3,000 people and left much of the city in ruins, while the section south of Parkfield, including the long stretch near the Salton Sea, produced the magnitude-7.9 Fort Tejon earthquake of 1857 and has been largely silent since.
A Segment That Hasn’t Moved Since Colonial Times
Paleoseismic trenches dug across the southernmost San Andreas, including sites near the Salton Sea and Coachella Valley, expose layers of sediment offset each time the ground ruptured in the past. Dating of that sediment indicates the last major rupture on this stretch occurred sometime around 1690, and USGS researchers describe the Southern San Andreas Fault as one of the most closely watched seismic hotspots in the world precisely because it has gone more than 300 years without a major rupture, well beyond what the segment’s own history suggests is typical. That gap is the main reason scientists describe the southern San Andreas as significantly overdue, even though “overdue” in earthquake science is a probability statement, not a due date.
Reading Centuries of Strain in Buried Sediment
Because instrumental seismographs have only existed since the early 20th century, scientists studying the southern San Andreas rely heavily on paleoseismology, the study of physical evidence of past earthquakes preserved in the ground. Trenches cut across the fault reveal buried stream channels, cracked soil horizons, and other features offset by a specific amount each time the fault ruptured, allowing researchers to count individual past earthquakes and estimate their approximate size and timing. Combined with GPS and satellite radar measurements tracking how much strain has accumulated across the fault today, this record gives scientists a physical basis for estimating stored energy along this section, even without a single modern instrumental recording of the last rupture.
What USGS Models Say Could Happen Next
Ongoing USGS research along the fault feeds into statewide earthquake forecasts, which currently estimate roughly a 60 percent probability of a magnitude 6.7 or larger earthquake occurring somewhere on the broader San Andreas fault system within the next 30 years. Scenario modeling of a full rupture of the long-locked southern segment, drawing on the accumulated strain estimated from paleoseismic and geodetic data, suggests a quake in the magnitude 7.8 range is plausible, with shaking strong enough to damage freeway overpasses, rupture buried water and gas lines, and affect tens of millions of people across Southern California’s densely populated valleys and coastal cities. Emergency planners have used scenarios like this, including the widely cited “ShakeOut” exercise developed with USGS scientists, to run statewide earthquake drills aimed at testing hospitals, utilities, and transportation networks against the kind of shaking such a rupture would produce.
Why “Overdue” Doesn’t Mean “Imminent”
Recent research into faults connected to the southern San Andreas, including a 2025 USGS study of faults beneath the Salton Sea, has examined how stress transfers between nearby fault segments. That research found that a dense network of smaller, vertically-moving faults beneath the Salton Sea has recorded at least four ancient displacement events that may have coincided with past San Andreas ruptures, and that computer models suggest the direction a future rupture travels, north-to-south versus the reverse, could determine whether it triggers additional movement on those buried faults. Findings like these do not change the underlying probability estimates, but they illustrate why scientists continue to expand monitoring networks along the fault, including seismic sensors that feed the USGS-operated ShakeAlert earthquake early warning system for the West Coast, rather than attempt to predict an exact date. An elevated probability is treated as a call for stronger building codes, retrofitted infrastructure, and public readiness, not a countdown to a specific day.
This article was produced with the assistance of AI and reviewed by Morning Overview editors.
More from Morning Overview
