Evolution of stress fields and faulting in seismic zones

Measurements indicate that stress magnitudes in the crust are normally limi ted by the frictional equilibrium on pre-existing, optimally oriented fault s. Fault zones where these limitations are frequently reached are referred to as seismic zones. Fault zones in the crust concentrate stresses because their material properties are different from those of the host rock. Most f ault zones are spatially relatively stable structures, however the associat ed seismicity in these zones is quite variable in space and time. Here we p ropose that this variability is attributable to stress-concentration zones that migrate and expand through the fault zone. We suggest that following a large earthquake and the associated stress relaxation, shear stresses of a magnitude sufficient to produce earthquakes occur only in those small part s of the seismic zone that, because of material properties and boundary con ditions, encourage concentration of shear stress. During the earthquake cyc le, the conditions for seismogenic fault slip migrate from these stress-con centration regions throughout the entire seismic zone. Thus, while the stre ss-concentration regions continue to produce small slips and small earthqua kes throughout the seismic cycle, the conditions for slip and earthquakes a re gradually reached in larger parts of, and eventually the whole, seismoge nic layer of the seismic zone. Prior to the propagation of an earthquake fr acture that gives rise to a large earthquake, the stress conditions in the zone along the whole potential rupture plane must be essentially similar. T his follows because if they were not, then, on entering crustal parts where the state of stress was unfavourable to this type of faulting, the Fault p ropagation would be arrested. The proposed necessary homogenisation of the stress field in a seismic zone as a precursor to large earthquakes implies that by monitoring the state of stress in a seismic zone, its large earthqu akes may possibly be forecasted. We test the model on data from Iceland and demonstrate that it broadly explains the historical, as well as the curren t, patterns of seismogenic faulting in the South Iceland Seismic Zone.