Group A, Poster #051, Seismology

Variability in finite-fault slip models of the 2011 Tohoku-Oki earthquake and implications for observationally constrained dynamic rupture simulations

Jeremy Wing Ching Wong, Wenyuan Fan, & Alice-Agnes Gabriel
Poster Image: 

Poster Presentation

2022 SCEC Annual Meeting, Poster #051, SCEC Contribution #12372 VIEW PDF
Megathrust earthquakes are among the most damaging natural hazards. Imaging their rupture processes can provide insights into mitigating the associated hazards. However, finite-fault models suffer large uncertainties which are often poorly quantified. For example, there exists a multitude of published kinematic models of the 2011 Mw 9.0 Tohoku-Oki earthquake, which differ in their imaged kinematics and lead to different interpretations (e.g., Lay, 2018). Here we compare 25 of these models to systematically quantify their variability and investigate their physical implications. The finite fault models of the Tohoku-Oki earthquake are constrained using a variety of datasets, including onshore ...and offshore geodetic measurements, regional and teleseismic seismic recordings, as well as tsunami and seafloor displacement measurements. The models are further obtained by employing varying parameterizations, regularizations, and optimization algorithms during the respective inversion. Specifically, the models differ in their capability of resolving potential multiple slip episodes and the northern extent of the slip distribution. The amplitude of the shallow, near-trench slip evolution, which has important implications for tsunami generation, is also poorly constrained. To quantitatively compare finite fault models, we design a uniform processing scheme to extract the coherent and incoherent slip features of the models. We first project the kinematic models to the slab 2.0 megathrust geometry and then upscale them to the same evenly-spaced grids, which are finally used for analyses of their statistical and physical variability. The projected finite-fault models will be used for forward computing synthetic observables and comparing their sensitivity to the identified statistical features. We will also select representative models to constrain 3D physics-based dynamic rupture simulations of the Tohoku-Oki earthquake (following Tinti et al., 2021). We aim to identify the key faulting conditions and the underlying mechanisms that governed the complex rupture propagation of the 2011 Tohoku-Oki Earthquake.