No scientist has ever predicted a major earthquake. Better forecasts would inform the decisions people and organizations make to reduce earthquake risk, and help communities prepare for destructive quakes.
What we know
We know where the risk is high, but not when a rupture will happen.
What blocks us
No reliable precursor signal has been found.
Current leads
AI on seismic data, fiber-optic cables used as sensors.
Sub-questions
Can the far more detailed earthquake catalogs built by AI reveal new patterns in how earthquakes happen?
Can combining AI with the physics of faults produce earthquake forecasts that are actually useful?
Do signals other than ground shaking, such as chemical or atmospheric changes, carry real warning of an earthquake?