Deformation processes in great subduction zone earthquake cycles

Crustal deformation associated with great subduction zone earthquakes yields important information on mantle rheology and slip evolution of the megathrust. We have used three-dimensional viscoelastic finite element models to study the contemporary crustal deformation of three margins, Sumatra, Chile, and Cascadia, that are presently at different stages of their great earthquake cycles. At Sumatra where an M_w 9.2 earthquake occurred in 2004, all the GPS stations are moving seaward. At Chile where an M_w 9.5 earthquake occurred in 1960, coast GPS stations are moving landward, obviously due to the re-locking of the fault, while the inland stations are still moving seaward. At Cascadia where an M_w 9.0 earthquake occurred in 1700, all the GPS stations are moving landward. The earthquake cycle deformation at Alaska where an M_w 9.2 earthquake occurred in 1964 is similar to that of Chile, and the deformation at NE Japan where an M_w 9.0 earthquake occurred in 2011 is similar to that of Sumatra. Model results indicate that the earthquake cycle deformation of different margins is governed by a common physical process. A great earthquake causes the upper plate to move towards the trench and induces shear stresses in the upper mantle. After the earthquake, the fault is re-locked, causing the upper plate to move landward. However, portions of the fault undergo aseismic afterslip for a short duration, causing the overriding areas to move seaward. At the same time, the viscoelastic stress relaxation of the upper mantle causes prolonged seaward motion in inland areas including the forearc and the back arc. After a long time when the earthquake-induced stresses have mostly relaxed, the upper plate moves landward due to the re-locking of the fault. The model of the 2004 Sumatra earthquake indicates that the afterslip must be at work immediately after the earthquake, and the characteristic time of the afterslip is ~1 yr. With the incorporation of the transient (biviscous) rheology, the model well explains the near-field and far-field postseismic deformation within a few years after the 2004 Sumatra event. For all the margins modeled, the steady-state (Maxwell) viscosity of the continental upper mantle is determined to be ~10¹⁹ Pa s, two orders of magnitude lower than that of the global value obtained through global postglacial rebound analyses. Based on the model for the 2004 Sumatra earthquake, the transient (Kelvin) viscosity of the continental mantle is one to two orders of magnitude lower than that of the stead-state viscosity. Long-term postseismic deformation is controlled mainly by the steady-state viscosity of the mantle and is relatively better understood. For the short-term postseismic deformation, the interaction of the afterslip of the fault and the transient deformation of the mantle is still poorly understood. Geodetic monitoring following the 2010 M_w 8.8 Maule earthquake and 2011 M_w 9.0 Tohoku earthquakes is expected to improve greatly our understanding the short-term deformation over the next few years.