Beneath the everyday routines of millions of Southern Californians, two tectonic plates are sliding past each other along a fault line that has not fully released its accumulated strain in generations. The San Andreas Fault does not announce its progress with tremors large enough to notice most days, but seismologists who study it describe a fault system that is steadily storing up energy for an eventual large rupture rather than gradually and harmlessly creeping. Understanding what scientists actually mean by “overdue” and “the Big One” requires separating documented probabilities from popular shorthand, since the two are often blurred together in casual conversation about California earthquakes. The distinction matters because the fault does not behave uniformly along its full length; some stretches slip gradually while others remain locked for generations, and it is specifically those locked stretches that concentrate the risk seismologists spend the most time studying.
What Makes the San Andreas Different From Other California Faults
The San Andreas Fault marks the boundary where the Pacific Plate and the North American Plate grind past one another, running for roughly 800 miles through California from the Salton Sea in the south to the state’s northern coast. Unlike faults that thrust one block of crust up and over another, the San Andreas is primarily a strike-slip fault, meaning the two sides move horizontally past each other, which is part of why a large rupture there tends to produce long, rolling ground shaking over a wide area rather than a single sharp jolt confined to one spot.
Not every segment of the fault behaves the same way. Some sections release their strain gradually through a slow, mostly harmless motion known as fault creep, easing pressure a little at a time without a major earthquake. Other segments, including long stretches of the southern San Andreas, are considered locked, meaning the two sides are stuck against each other by friction even as the plates beneath them keep moving, which lets stress accumulate steadily until it is finally released all at once in a large rupture rather than bleeding off gradually.
The Odds According to the USGS: Roughly a Coin Flip Every 30 Years
Government seismologists have put real numbers on the risk rather than leaving it as vague menace. A U.S. Geological Survey analysis found that eight earthquakes of magnitude 7 or larger struck Southern California over a 200-year stretch, and treating those events as a statistical process yields a 67 percent probability, plus or minus 23 percentage points, of at least one similar earthquake occurring somewhere in Southern California within any given 30-year window. Five of those eight historical earthquakes struck specifically on the San Andreas fault system, which is why the fault draws so much more scientific attention than the region’s many smaller, less active faults.
What a Magnitude-7.8 Rupture Would Actually Look Like
The specific scenario behind “the Big One” comes from a detailed USGS-led study of a hypothetical magnitude 7.8 earthquake on the southern San Andreas Fault, led by USGS seismologist Lucy Jones with more than 300 contributing scientists and engineers. According to the resulting ShakeOut Scenario, an earthquake of that size would produce shaking lasting nearly two minutes, hit hardest in the Coachella Valley, the Inland Empire and the Antelope Valley near the fault itself, and create additional pockets of intense shaking farther away in the San Gabriel Valley and East Los Angeles, where soft sediment traps and amplifies seismic waves. The study estimated the event could cause more than 1,800 deaths, roughly 50,000 injuries and around $200 billion in damage and other economic losses, alongside severe, long-lasting disruption to water pipelines, highways and other infrastructure that crosses the fault at multiple points. The scenario has since been used as the basis for statewide emergency-response exercises, precisely because its scale, more than 5,000 times stronger than the magnitude 5.4 earthquake that struck Southern California the same year the scenario was published, is difficult to plan for using the assumptions built into most everyday disaster drills.
How the Newest Forecast Model Changed the Picture
Earthquake forecasting itself has evolved since those early probability estimates were first calculated. The USGS’s third-generation statewide model, known as UCERF3, incorporated new data and lessons from recent earthquakes to reduce the estimated likelihood of moderate-sized events in the magnitude 6.5 to 7.5 range while increasing the estimated likelihood of larger earthquakes. The change reflects a key modeling improvement: earlier forecasts largely treated faults as separate, isolated segments, while UCERF3 accounts for multifault ruptures, in which a single earthquake can break across more than one fault at once, a scenario that tends to produce a larger event than older models assumed.
How Scientists Track the Fault’s Slow Build-Up
Because the buildup happens too slowly to feel, researchers rely on networks of GPS stations anchored to bedrock on either side of the fault, tracking how many millimeters the ground on each side shifts relative to the other every year. Those measurements, combined with satellite radar that can detect ground deformation across a wide area between visits, let scientists map exactly which segments are creeping harmlessly and which remain locked and accumulating stress. The same monitoring networks also record the fault’s much more frequent small earthquakes, most too weak to notice, which researchers use to refine estimates of where locked segments begin and end rather than relying solely on the historical record of past large ruptures.
Why “Overdue” Doesn’t Mean “Imminent”
None of this amounts to a prediction that the fault will rupture on any particular day, and seismologists are careful to distinguish long-term probability from short-term forecasting. Strain along a fault segment builds continuously and invisibly, without a clock counting down to a specific date, so a segment can be statistically “overdue” for a large earthquake relative to its historical average recurrence interval and still go years or decades longer without breaking. Earthquake science has never produced a reliable method for predicting the exact day, week or even year a specific fault will rupture, which is why agencies frame the risk in probabilities over decades rather than countdowns, and why “overdue” is a statement about historical averages rather than a warning of an event about to happen tomorrow.
What the accumulated research does establish clearly is the scale of the eventual event once it happens: a fault capable of producing multifault ruptures, in a region where tens of millions of people now live directly along or near its path, in structures and infrastructure that were mostly not designed with the newest, larger-event forecasts in mind. That gap between an aging built environment and an updated hazard model is a large part of why California building codes, retrofit programs and emergency planning keep getting revised even without any single earthquake forcing the change, since the risk itself is understood to be shifting upward as the science around it improves. Older buildings, bridges and pipelines built before the newest forecasts existed were engineered to standards that assumed a smaller worst-case event than researchers now consider plausible, which is exactly the kind of mismatch retrofit programs are meant to close before, rather than after, a major rupture occurs.
This article was produced with the assistance of AI and reviewed by Morning Overview editors.
More from Morning Overview
