Modeling of Earthquake Ground Motion: From the Source to the Site.
Loading...
Date
2024-08-07
Authors
Nye, Tara
Journal Title
Journal ISSN
Volume Title
Publisher
University of Oregon
Abstract
This dissertation investigates different approaches for improving modeling ofearthquake ground motion by focusing efforts on the different physical properties which
are traditionally thought to contribute to shaking–– the source, the path, and the site. The
robustness and informativeness of earthquake simulations and empirically-derived ground
motion models rely on accurate modeling of these different components. Methods to
improve ground motion modeling are highly studied; however, due to the complexity
of processes that contribute to shaking intensities and uncertainty, much is still left
poorly understood, especially when modeling rare or atypical events. This dissertation
utilizes both simulated and real earthquake data to better constrain source, path, and site
parameters.
In this dissertation, we use a one-dimensional semistochastic model to simulate
data for the 2010M7.8 Mentawai tsunami earthquake, as well as for 112 moderate-to-large
arbitrary Cascadia Subduction Zone (CSZ) events. Tsunami earthquakes and CSZ ruptures
are scarce, which results in minimal data to analyze and thus a poorer understanding of
the rupture physics. However, a future occurrence of either scenario is likely to result in
significant damage. With the paucity of real data, simulations are necessary to train and test
earthquake and tsunami early warning algorithms. We use near-field observed data from
the Mentawai tsunami earthquake to ascertain source parameter values needed to simulate
tsunami earthquake-type events. With the absence of recordings from a CSZ interface
event, we use global ground motion and scaling models to validate and tune modeling of
the source and path in the simulation model. Through our analyses, we find the simulations
to be well-modeled and representative of these uncommon events.
Finally, we constrain site effects in the seismically-active San Francisco Bay
Area (SFBA) by using real recordings of small-to-moderate earthquakes to estimate the
high frequency attenuation parameter (κ) and its site contribution (κ0). We develop an
automated algorithm for selecting the frequency bounds used to estimate κ, and we find
spatial trends in κ0 to be consistent with regional anelastic attenuation models and shallow
geology. Through ground motion analyses, we find κ0 to reasonably model regional ground
motion and is thus a valuable contribution to future ground motion studies in the SFBA.
This dissertation includes previously published and unpublished co-authored
material.
Description
Keywords
Earthquake Simulations, Ground-Motion Modeling, Seismology