Proving Fault Geometry with Finite-Fault Inversion of Teleseismic Data

An attempt of finite-fault modeling with five-basis components of double-couple moment tensor has successfully extracted information on fault geometry included in teleseismic data. In that framework of inversion, a source model is represented as spatiotemporal distribution of potency-density tensor, and geometric feature of fault system can be extracted from the spatial variation of potency-density tensors. However, location of resultant potency-density tensor on an assumed model plane, which is adopted by modeler, can be deviated from a true fault plane, and hence it is difficult to directly interpret the potency-density tensor distribution as a fault geometry. In this study, we propose a way of constructing fault geometry by sole use of teleseismic data with an iterative process of finite-fault inversion resolving potency-density tensor, which is efficient for directly evaluating relationship between rupture propagation and a geometric complexity of fault system. We construct a fault geometry with following steps: (1) estimating potency-density tensor distribution along a (N-1)th model plane by performing the finite-fault inversion, (2) updating an Nth non-planar model geometry whose spatial gradients are consistent with strike and dip angles of the potency-density tensor distribution, and (3) doing the finite-fault inversion to estimate potency-density tensor distribution along the Nth model plane again until difference of geometry between the (N-1)th and Nth model planes becomes acceptably small. We applied this approach to recent large earthquakes which are thought to have occurred along a geometrically complex fault system, and found that geometry of the reconstructed model planes are comparable to the ones obtained by high resolution geodetic measurements.