← Back to Sessions

Sensitivity Kernels for Basin Path and Site Effects

Session: Numerical Modeling of Earthquake Motion, Rupture Dynamics, Seismic Noise, Wave Propagation and Inverse Problems I

Type: Oral

Date: 4/23/2021

Presentation Time: 09:45 AM Pacific

Description: 

Numerous studies have explored how shallow wave propagation can amplify and focus seismic waves, significantly increasing seismic hazard. In complex basin structures, seismologists often must rely on simulations or careful timeseries analysis to determine whether the strongest shaking resulted from direct S-wave amplifications, from basin or topographic focusing, from converted surface waves at a basin’s edge, etc.

In this work we explore a systematic method for characterizing which part of a sedimentary basin or other geologic structure was most responsible for the observed amplitudes. The use of adjoint, finite-frequency based sensitivity kernels is well developed for tomography applications but seldom applied to hazard. In tomographic applications, an adjoint kernel is generally used to characterize the region of a model which would need to be altered to reduce some misfit between observations and synthetics. Here we ask the question: what region of the model most significantly contributed to an observed high-amplitude signal?

We note such methods have been explored for source and path sensitivities hazard by Day, Roten and Olsen (2012), but while that work provided a detailed and thorough exploration of a large-scale Shakeout scenario, our approach in this work will be of smaller scale and systematic. By examining the kernels from numerous events and by slowly adding geologic complexity to the basin shape, we hope to shed light on when and how different wavetypes are excited or when they are important contributors to hazard.

Presenting Author: Daniel Bowden

Student Presenter: No

Authors