The Strain Before the Snap
The logic behind the seismic gap theory is based on a fundamental concept called elastic rebound. Following the great 1906 San Francisco earthquake, geologist Harry Fielding Reid studied how the ground had deformed. He realized the Earth's crust isn't perfectly rigid. As tectonic plates grind past each other—at speeds of 1 to 10 centimeters per year—sections of a fault can get locked by friction. While the plates keep moving, the rock around the locked fault bends and deforms, storing elastic energy like a stretched rubber band. When the accumulated stress finally overcomes the friction, the fault ruptures. The rock snaps back to a less strained position, releasing the stored energy as seismic waves.
In the 1970s, seismologists took this idea further. They proposed that long, active faults rupture in segments. If historical records showed that segments on either side of a particular "quiet" section had ruptured more recently, that quiet section, or "seismic gap," was the most likely candidate for a future earthquake. The longer the gap remained quiet, the more strain it was presumably accumulating, and the larger the potential earthquake. The coordinates for this location (47.5, 155.3) point to the Kuril-Kamchatka Trench, a hotbed of seismic activity and an important area where this theory was first developed.
Successes and spectacular failures
The seismic gap model has had some notable successes. A segment of the San Andreas Fault in California was identified as a seismic gap, and it subsequently ruptured in the 1989 magnitude 6.9 Loma Prieta earthquake. The Kuril Islands arc, a subduction zone where the Pacific Plate dives beneath the Okhotsk Plate, was a clear example. A 500-kilometer-long gap that had not seen a major event since 1780 was identified. In November 2006, a magnitude 8.3 earthquake ruptured the southern half of this gap.
However, the theory's predictive power is far from perfect. The Parkfield experiment, initiated in 1985, was designed to capture a predicted magnitude 6 earthquake on a segment of the San Andreas Fault that had ruptured with striking regularity. Based on a 22-year average interval, the quake was expected before 1993. It did not occur until 2004, and when it did, it lacked some of the anticipated precursor signals.
The most dramatic failure of the simple gap model was the 2011 Tōhoku earthquake in Japan. While the region was known to be seismically active, no one anticipated a magnitude 9.0–9.1 event. The earthquake ruptured a much larger area than expected, cascading through multiple segments that were not considered mature seismic gaps. The slip on the fault was immense, reaching up to 50 or 60 meters in some places, far greater than what the accumulated strain over the known quiet period could explain. This event demonstrated that faults can rupture in complex ways that defy simple segmentation models. As a result, many seismologists now view the seismic gap concept as a useful, but overly simplified, tool for assessing earthquake hazard.