Well before the plate tectonic theory was introduced, Reid suggested that the forces that cause an earthquake are not near the earthquake, but very far away.

Over hundreds to thousands of years, these distant forces cause a gradual build up of elastic energy.  The diagram below shows how they have distorted the ground compared to a recently built fence.

The ground continues to slowly distort until an existing weakness in the Earth (a fault) cannot take the strain. Then within a short period of time (usually a matter of seconds) elastic energy is suddenly and catastrophically released, resulting in an earthquake, and the ground moves as a result.

The distortion of the fence is what Reid observed from the survey data taken just before and after the San Francisco earthquake.

Over long periods this repeated cycle results in large movements along the line of the fault (offsets).

But Reid’s elastic-rebound model leaves many questions unanswered.  For example, it ignores changes in the properties of rocks with depth.

There is broad agreement that most continental earthquakes occur in the seismogenic crust – the upper 10 to 20 km – and that this behaves elastically, just as in Reid’s model. Below this, where rocks are hotter, the material properties and behaviour of continental crust are still uncertain. This is something that we are investigating.