Time Dependence of the Stress Shadowing Effect and Its Relation to the Structure of the Lower Crust

S. J. Kenner and P. Segall, Stanford University

Abstract

Great earthquakes, like the 1906 San Francisco earthquake, perturb the regional stress field and can generate “stress shadows” that delay the occurrence of subsequent M ³ 5.5 events. We investigate how time dependent postseismic stress transfer between the three subparallel strike-slip faults in northern California - the San Andreas, Hayward, and Calaveras faults, and their northern extensions - depends on the presence and geometry of lower crustal shear zones. Models incorporating Maxwell viscoelastic relaxation of the mantle, a lower crustal detachment surface, and vertical shear zones beneath the seismogenic faults are compared. Short-term relaxation of lower crustal shear zones enhances the amplitude and lateral extent of the stress shadow. Conversely, longer period mantle relaxation reloads the entire crust. As a consequence, after a 1906 type event on the San Andreas fault, the stress shadow on the Hayward fault will be greater in magnitude and duration if the faults are connected via lower crustal structures with time dependent rheologies. In addition, our results demonstrate that, for certain geometries, the total stress (tectonic plus postseismic perturbation) may actually continue to decrease during the decades immediately following a great earthquake. This effect has not been documented in prior time dependent models of stress accumulation during the earthquake cycle.

Figure 1: Finite element mesh for (a) the crustal fault model, which includes only coseismic fault, (b) the shear zone model, (c) the free slip model, and (d) the lower crustal detachment model. Black box in (a) defines area shown in (b), (c), and (d). The Hayward fault (H.F.) and Calaveras fault (C.F.) are located 32 km and 44 km east of San Andreas fault.

Figure 2: Evolution with time of the fault parallel shear stress perturbation at ~7 km depth after a 1906 type event for (a) the San Andreas fault and (b) the Hayward fault.

Figure 3: For the crustal fault model in Figure 1, (a) contours of the postseismic shear stress perturbation with depth ~90 years after coseismic rupture and (b) the evolution of fault parallel shear stress perturbation through time at depth of 7.2 km. In (b), increasing stresses denote reloading. Decreasing stresses indicate continued relaxation. Coseismic stress change is given by time = 0 year curve.

Figure 4: For the shear zone model in Figure 1, (a) contours of the postseismic shear stress perturbation with depth ~90 years after coseismic rupture and (b) the evolution of fault parallel shear stress perturbation through time at depth of 7.2 km. In (b), increasing stresses denote reloading. Decreasing stresses indicate continued relaxation. Coseismic stress change is given by time = 0 year curve.

Figure 5: For the freely slipping model in Figure 1, (a) contours of the postseismic shear stress perturbation with depth ~90 years after coseismic rupture and (b) the evolution of fault parallel shear stress perturbation through time at depth of 7.2 km. In (b), increasing stresses denote reloading. Decreasing stresses indicate continued relaxation. Coseismic stress change is given by time = 0 year curve.

Figure 6: For the detachment model in Figure 1, (a) contours of the postseismic shear stress perturbation with depth ~90 years after coseismic rupture and (b) the evolution of fault parallel shear stress perturbation through time at depth of 7.2 km. In (b), increasing stresses denote reloading. Decreasing stresses indicate continued relaxation. Coseismic stress change is given by time = 0 year curve.

Reference: Kenner, S.J., and P. Segall, Time Dependence of the Stress Shadowing Effect and Its Relation to the Structure of the Lower Crust, Geology, v. 27, p. 119-122, February 1999.