The San Andreas Is Under More Stress Than at Any Time in 1,000 Years

A team at the University of Bern, with Scripps and USGS co-authors, built a physics-based model of the southern San Andreas and San Jacinto faults and ran it through 1,000 years of earthquakes reconstructed from trenches, radiocarbon dates and historical records. The result: tectonic stress on parts of the system has reached — and in places passed — the highest values anywhere in that thousand-year run. One segment near San Bernardino sits at 3.6 megapascals, above its previous peak of nearly fifty years ago.

The interesting part is Cajon Pass, where the two faults meet. The model shows it works as an earthquake gate: a rupture stops at the junction when the two sides carry different loads, and runs through when they are similarly stressed. Both sides are similarly stressed now. The authors are explicit that this is not a prediction of when — it is a statement about how loaded the spring is.

This is the southern end of the fault. My station in Soquel sits at the northern end, about six miles from the 1989 Loma Prieta epicenter, watching the San Andreas and the Calaveras every day. In July I wrote about new modeling that a Cascadia rupture could load the northern San Andreas. Read the two together and the picture is a fault loaded at both ends.

Source: Burkhard, Smith-Konter, Scharer & Sandwell, Journal of Geophysical Research: Solid Earth 131(6), 2026, doi:10.1029/2025JB033213; summarized by ScienceDaily, October 1, 2026. Read the original for the specifics.