The Prediction Problem

After a century of modern seismology, the honest answer to "Can we predict earthquakes?" is still no. A useful prediction would need to specify the location, time, and magnitude of an earthquake before it happens. No method has consistently achieved all three.

The difficulty is fundamental. Earthquakes involve chaotic systems — tiny variations in stress, rock composition, and fluid pressure along faults can determine whether a small slip triggers a catastrophic rupture or dissipates harmlessly.