AI Uncovers 92 Hidden San Andreas Slow-Slip Events Linked to 500,000 Low-Frequency Quakes
Updated
Updated · Gizmodo · Jul 10
AI Uncovers 92 Hidden San Andreas Slow-Slip Events Linked to 500,000 Low-Frequency Quakes
2 articles · Updated · Gizmodo · Jul 10
Summary
Eight years of borehole strain data from four Parkfield sensors let researchers identify 92 previously unknown slow-slip events along the San Andreas Fault, a signal conventional methods had missed.
The team found those aseismic slips often coincided with low-frequency earthquakes occurring within 10 kilometers and shallower than 20 kilometers, suggesting slow sliding helps set local stress conditions.
Machine learning was key because the subtle deformations can last from seconds to weeks and sit in a measurement gap between seismometers and high-precision GPS.
Researchers say the result does not amount to earthquake prediction, but it could sharpen understanding of fault behavior and possible early warning signs for larger quakes.
The group plans to test the approach on other faults to verify the link, noting the Parkfield study compared roughly 500,000 low-frequency quakes with only 92 detected slow-slip events.
If AI can hear the quiet whispers before a quake, can it tell which ones are deadly threats?
As warning systems already trigger false alarms, could AI's new precursor alerts create even greater public panic?
92 Hidden Slow Slip Events Detected by AI on San Andreas Fault: New Insights for Earthquake Risk
Overview
A pioneering team led by Dr. Zahra Zali used artificial intelligence to uncover 92 previously hidden short-duration slow slip events beneath the Parkfield section of the San Andreas Fault. These silent slips, which release stress over hours to days without noticeable shaking, had long escaped standard monitoring and created a critical blind spot in understanding fault behavior. By applying advanced deep learning to borehole strainmeter data, the team revealed these subtle movements, offering new insights into the complex mechanics of the fault and marking a breakthrough in seismic monitoring and earthquake science.