Stressing of locked patches along a creeping fault

Abstract

An analysis is presented of the stressing of locked patches along a fault zone which is creeping elsewhere. The model consists of strike-slip faulted elastic lithospheric plates loaded in a manner equivalent to imposition of a remotely uniform tectonic shear stress. The fault zone has an inhomogeneous strength distribution both depth-wise and along strike. It is modeled as being composed of locked patches and freely slipping parts treated as cracks. The solution is given with the use of the “line-spring” model which analyzes the problem by thickness-averaged plane stress theory for lithospheric plates which slip along a discontinuity cut at the plate boundary. As a boundary condition, thickness-averaged stress and slip at each local section along the cut was related to one another by the result of an antiplane strain analysis of slip for the crack or crack pair which describes the slipping and locked depth ranges at that section. The analysis indicates that the slip distribution along creeping parts of the fault, as well as the stress distribution along locked patches, depends strongly on the geometry of these zones. The model is used to examine stress concentrations associated with a slip-deficient seismic gap along strike and to study the effect of local irregularities in the margin of a locked region. It is also used to simulate slip and stressing processes associated with the creeping portion of the San Andreas fault in central California, between the presently locked zones of the great 1906 and 1857 ruptures, and to constrain the nature of an apparently locked zone at the southeastern end which ruptures in characteristic Parkfield earthquakes. Near-fault creep and broadscale displacement data along the fault since the 1966 Parkfield earthquake and inferences from seismicity distributions are used. Limitations of the modeling procedure at short spatial wavelengths prohibit an accurate description of the Parkfield locked patch, but results suggest that it may be localized and occupy a small fraction of area of the normal seismogenic zone. An effective remote stressing rate of order 0.3 × 10−6 × shear modulus/yr is inferred, together with a less well constrained 30 to 40 km lithospheric thickness, for consistency with the displacement data. Results enable estimates of stress accumulation along the locked 1857 rupture zone and the build-up of fracture energy release capability (of order 107 J/m2 in 150 yr) at its lower margin.