Spherical‐Earth finite element model of short‐term postseismic deformation following the 2004 Sumatra earthquake

yan hu,kelin wang
DOI: https://doi.org/10.1029/2012JB009153
2012-01-01
Abstract:Decadal-scale postseismic deformation of subduction earthquakes has been widely modeled using a Maxwell viscoelastic Earth with a mantle-wedge viscosity of about 10(19) Pa s. Short-term postseismic deformation within a few years after the earthquake remains a more challenging problem because of the predominance of afterslip of the megathrust and the potentially more complex mantle rheology. Questions include the necessity of invoking the transient rheology and the relative importance of contributions from afterslip and viscoelastic relaxation. In this work, we address these questions by developing a spherical-Earth viscoelastic finite element model for the short-term postseismic deformation following the magnitude 9.2 Sumatra earthquake of 2004. The model consists of elastic overriding and subducting plates and a viscoelastic mantle of bi-viscous Burgers rheology, with the continental mantle (mantle wedge) being less viscous than the oceanic mantle by a factor of ten. Primary observational constraints for the short-term postseismic deformation include similar to 1 year net displacements of nine near-field GPS sites and similar to 3 year time series from three far-field sites several hundred kilometers from the 2004 rupture zone. Model results indicate that the afterslip of the fault must be at work within the first few years after the earthquake. The observed deformation is best explained with a model that includes both the afterslip and transient rheology. In the preferred model, the transient and steady state viscosities of the mantle wedge are 5 x 10(17) Pa s and 10(19) Pa s, respectively.
What problem does this paper attempt to address?