The release of accumulated stress at a particular point during an earthquake alters the shear and normal stresses along the tectonic plate boundary and surrounding fault lines.
By measuring and calculating the stress transfer following seismic activity, it is possible to subsequently construct basic contour maps of regions where there have been positive stress changes and are therefore of higher risk of being potential epicenters of future large earthquakes. Calculations have revealed that when an earthquake occurs, approximately 80% of the energy is released as seismic waves, whereas the remaining 20% is stored and transferred to different locations along the fault, making those specific regions more susceptible to future earthquakes.
Predicting earthquakes by using the theory of stress transfer has important potential applications. The main rival technique for forecasting, the statistical analysis of patterns in seismic activity, suffers from a contradiction. Foreshocks are deemed evidence of the potential for a future high-magnitude earthquake, but the lack of foreshocks along faults known to be active has been considered an equally plausible potential precursor for large events.
The stress transfer theory has been used to predict the location of a magnitude 7.4 earthquake that occurred two years later in the port city of Izmit, Turkey, killing more than 30,000 people. A limitation of the theory as currently applied, due to insufficient understanding of plate kinematics, is that refining predictions with temporal constraints appears to be far more problematic; the team that gave the Izmit prediction had forecast an event near the city within thirty years.