A U.S. Geological Survey page lists a 31% probability for a magnitude 7.5 earthquake in the Los Angeles region over a 30-year period. The figure comes from the Uniform California Earthquake Rupture Forecast 3, or UCERF3. It is a model result from a forecast published in 2015, not a live 30-year countdown starting today.
That distinction does not make the risk trivial. It makes the claim honest. Earthquake forecasts describe probabilities across regions and time windows; they cannot say when a specific earthquake will occur. Preparing for strong shaking is practical, while treating a model percentage as a calendar prediction is not.
The 31% result has a specific source
The USGS Los Angeles probability FAQ says UCERF3 gave the region a 31% probability of a magnitude 7.5 earthquake within the model’s next 30 years. It also lists 60% for magnitude 6.7 and 46% for magnitude 7 in the same region.
The wording “according to UCERF3” is essential. The number is not a statement that an earthquake is 31% likely in the next 30 years beginning September 2026. It preserves the study’s specified forecast window and the model on which the calculation was based.
USGS published the underlying time-dependent probability study in 2015. The publication record identifies it as the Working Group on California Earthquake Probabilities’ UCERF3 analysis and notes that the model’s use should be evaluated case by case.
A probability model is not an earthquake prediction
An earthquake prediction would specify a particular time, location and magnitude before the event. USGS does not make those predictions. A long-term forecast instead combines fault data, historical intervals, possible rupture paths and uncertainty to estimate how often events of different sizes may occur over a broad period.
The UCERF3 fact sheet explains that the model incorporated multifault ruptures, allowing some scenarios to cross faults rather than stopping at assumed segments. That scientific improvement changed estimates for different parts of California; it did not reveal a date for the next Los Angeles earthquake.
The Los Angeles region contains many potential earthquake sources, so a regional percentage should not be read as a probability for one named fault or one neighborhood. Local ground conditions, building type and distance from a rupture all affect the shaking and damage a community might experience.
Why the model changed earlier estimates
In its 2015 news release, USGS said UCERF3 used new data and methods to revise earthquake likelihoods across California. It reported that the modeled likelihood of a magnitude 8 or larger earthquake statewide over 30 years increased compared with the previous assessment, in part because multifault ruptures were represented.
Another USGS hazard-model publication describes UCERF3 as an earthquake-rate model, paired with ground-motion models to estimate shaking hazards. That division matters: rupture likelihood and the expected shaking at a given site are related but not identical questions.
Scientists revise these models when they have new data, revised fault rates or improved methods. A responsible article therefore identifies the version and publication date instead of silently converting an older study into an evergreen present-tense forecast.
Preparation does not require a countdown
The practical message survives the correction. Californians live with a real long-term earthquake hazard, and households can reduce risk without waiting for a prediction. Ready.gov’s earthquake guidance recommends securing heavy furniture, making an emergency plan and knowing how to Drop, Cover and Hold On.
The 31% number remains meaningful when described correctly: it was a UCERF3 model estimate for the Los Angeles region over its stated 30-year horizon. It is not evidence that a new 30-year clock began today. Separating those ideas respects both the science and the people who use hazard information to prepare.
It also avoids a common error in earthquake stories: using the word “overdue” as if a fault keeps a public calendar. Faults accumulate strain and have histories, but recurrence intervals are not appointments. A probability forecast expresses uncertainty in a mathematically defined way; it does not convert uncertainty into a deadline.
For a reader, that means the best response to a big percentage is not to wait for a warning that will not come. It is to know where to take cover, make a plan with family members, and understand the local hazards of the places they use every day. Those choices remain useful whether a revised model raises, lowers or leaves unchanged a regional estimate.
UCERF3 is one version of a long-running California forecasting effort, not a promise that conditions will remain fixed for the next three decades. Its probabilities depend on the assumptions, fault data and rupture combinations available when the model was produced. A later scientific update could revise the numbers without making the earlier estimate fraudulent or meaningless.
Earthquake losses also vary widely within a metropolitan region. Soft soils can amplify shaking, older building types can be more vulnerable, and a nearby rupture can create effects unlike those at a more distant address. That is why a regional 31% result cannot substitute for building information, local hazard maps or emergency planning.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
More from Morning Overview
- The FBI tells phone owners to delete these toll-payment texts draining accounts nationwide
- Herbal supplements are landing Americans in the hospital with liver damage, doctors warn
- Long-term use of common heartburn pills is linked to kidney and dementia risk
- 9 pickup trucks with a reputation for falling apart after 100,000 miles
