2:30 PM - 2:45 PM
[SSS05-04] Effects of curvature and lateral heterogeneities in elasticity constants on coseismic deformation
Keywords:crustal deformation, lateral heterogeneity, finite element method, gravity, satellite geodesy, seafloor crustal deformation
Recently, new geodetic observation data such as seafloor crustal deformation data and satellite gravity data have been used to study wide-area crustal movements and variations in the gravity field, associated with great earthquakes. These data are compared with a theoretical model to infer a space-time distribution of the fault slip and the locked region and mantle rheology that explain the data.
The essential component to construct a theoretical model is a Green’s function (GF) prescribing the elastic response to an internal dislocation. Many existing GFs assume a flat Earth and treat the effects of self-gravitation only approximately. Such GFs cause modeling errors when dealing with wide-area deformations obtained by the above observations. A spherical Earth model can be used to calculate the global deformation including the change in the gravity field naturally. However, most models considering lateral heterogeneities, based on an ordinary finite-element method (FEM), do not solve the deformation over the entire sphere and the effects of self-gravitation are also approximated.
In this presentation, we first compare a flat-Earth model with the PREM in the coseismic elastic deformation due to a great earthquake and show that a spherical model is necessary to deal with near-field (i.e., seafloor deformation) and far-field (i.e., GNSS) observation data consistently. Next, we introduce a spectral FEM, which combines a 1-D finite element and a spherical harmonic representation, to calculate global coseismic deformation effectively. We modify this method to include lateral heterogeneities in the elasticity constants and apply it to a deep earthquake. The effects of lateral heterogeneities in Lame’s constants are evaluated by comparing a surface displacement obtained for a laterally heterogeneous model with that obtained for a spherically symmetric case (PREM). A preliminary result indicates that the effect of lateral heterogeneity in the shear modulus is larger than in the bulk modulus and that a local heterogeneity including the source region can affect the coseismic deformation as strongly as a regional-scale heterogeneity outside the source region.
The essential component to construct a theoretical model is a Green’s function (GF) prescribing the elastic response to an internal dislocation. Many existing GFs assume a flat Earth and treat the effects of self-gravitation only approximately. Such GFs cause modeling errors when dealing with wide-area deformations obtained by the above observations. A spherical Earth model can be used to calculate the global deformation including the change in the gravity field naturally. However, most models considering lateral heterogeneities, based on an ordinary finite-element method (FEM), do not solve the deformation over the entire sphere and the effects of self-gravitation are also approximated.
In this presentation, we first compare a flat-Earth model with the PREM in the coseismic elastic deformation due to a great earthquake and show that a spherical model is necessary to deal with near-field (i.e., seafloor deformation) and far-field (i.e., GNSS) observation data consistently. Next, we introduce a spectral FEM, which combines a 1-D finite element and a spherical harmonic representation, to calculate global coseismic deformation effectively. We modify this method to include lateral heterogeneities in the elasticity constants and apply it to a deep earthquake. The effects of lateral heterogeneities in Lame’s constants are evaluated by comparing a surface displacement obtained for a laterally heterogeneous model with that obtained for a spherically symmetric case (PREM). A preliminary result indicates that the effect of lateral heterogeneity in the shear modulus is larger than in the bulk modulus and that a local heterogeneity including the source region can affect the coseismic deformation as strongly as a regional-scale heterogeneity outside the source region.