Toward Dynamic Models of Contemporary Plate Boundary Deformation with Application to the Taiwan Collision Zone

点击翻译按钮获取中文摘要

The recent catastrophic earthquakes in Sumatra and Pakistan underscore the importance of developing techniques to identify potential large seismic sources and gaining an improved understanding of plate-boundary deformation processes. Crustal deformation models constrained by geodetic, seismic, paleoseismic, and geomorphic data are the appropriate tools for probing into the physics of plate-boundary deformation. The purpose of this research is to develop dynamic models of plate-boundary deformation that incorporate GPS measurements of crustal motions with a wide variety of geologic and geophysical data sets from the Taiwan arc-continent collision zone, one of the most densely instrumented, yet poorly understood, plate boundaries. The models are dynamic in the sense that slip on faults is governed by an assumed constitutive law that relates stresses in the lithosphere to fault slip. The models being developed in this research are an extension of related kinematic models in which slip on faults is prescribed based solely on the motions of lithospheric blocks, without regard to the driving forces. The overarching questions that motivate this work are directed toward an improved understanding of factors that control the spatial and temporal distribution of lithospheric deformation in Taiwan and other plate-boundary settings: 1) How is the rapid arc-continent convergence in Taiwan distributed among the active faults and folds? 2) What role do creeping and locked fault segments play in distributing this deformation?, 3) What role does fault geometry play in distributing the deformation?, 3) Where are future earthquakes likely to occur?, and 4) How can the various disparate data sets be integrated into a comprehensive 3D model of this ongoing deformation process? In particular, the research aims to integrate the dynamic model with various data sets to obtain estimates of fault slip rates in Taiwan and determine the spatial distribution of locked sections of active faults that are likely to rupture in future earthquakes. Model predictions of fault slip rates from GPS data are constrained by paleoseismic, geologic, and geomorphic estimates of long-term slip rates. Active fault geometry and locations of creeping and locked fault patches are constrained by relocated seismicity, distributions of earthquake focal mechanisms, and locations of small repeating earthquakes. Seismic tomography constrains crustal geometry and rheology.