Group C, Poster #234, Ground Motions
Investigating the Basin Effect on Ground Motions Using Laboratory Seismic Data Based on 3D-Printed Los Angeles Basin Structure
Poster Image:
Poster Presentation
2022 SCEC Annual Meeting, Poster #234, SCEC Contribution #12329 VIEW PDF
imulations utilizing 3D-printing technique. They built a physical model representing about 50-km long section of the Los Angeles Basin structure including the details (~100-m scale) of the basin interfaces. It was found that the high-frequency waves (>2 Hz) are selectively reflected at the basin edges, resulting in the reduction of high-frequency ground motions within the basin.
In this study, we examine additional datasets with different source locations—e.g., inside the basin—to further investigate the basin effect. In general, we find the reduction of the high-frequency ground motion within the basin is a common phenomenon regardless of the source locations. Considering parts of signals recorded in the laboratory do not exist in real seismograms (e.g., waves reflected at the physical model edge), we also develop a data filtering approach based on curvelet transform that can serve as a standard workflow for lab-experiment dataset preprocessing. The preprocessing allows us to obtain robust estimates of peak ground displacements, peak ground velocities, and power spectral densities (PGDs, PGVs, and PSDs), which are otherwise often overestimated due to interferences of undesired signals such as reflected seismic waves at model edges. Furthermore, by analyzing PSDs of filtered data, we make observations of previously undetected seismic phases that are reflected at which the basin depth changes sharply (by ~450 m). Our results demonstrate that high-frequency ground motions are significantly affected by small-scale—i.e., sub-wavelength—basin structures, including the edges and the geometry of the interfaces.
*Park et al., (2022), Scientific Report, https://doi.org/10.1038/s41598-022-08732-w
SHOW MORE
In this study, we examine additional datasets with different source locations—e.g., inside the basin—to further investigate the basin effect. In general, we find the reduction of the high-frequency ground motion within the basin is a common phenomenon regardless of the source locations. Considering parts of signals recorded in the laboratory do not exist in real seismograms (e.g., waves reflected at the physical model edge), we also develop a data filtering approach based on curvelet transform that can serve as a standard workflow for lab-experiment dataset preprocessing. The preprocessing allows us to obtain robust estimates of peak ground displacements, peak ground velocities, and power spectral densities (PGDs, PGVs, and PSDs), which are otherwise often overestimated due to interferences of undesired signals such as reflected seismic waves at model edges. Furthermore, by analyzing PSDs of filtered data, we make observations of previously undetected seismic phases that are reflected at which the basin depth changes sharply (by ~450 m). Our results demonstrate that high-frequency ground motions are significantly affected by small-scale—i.e., sub-wavelength—basin structures, including the edges and the geometry of the interfaces.
*Park et al., (2022), Scientific Report, https://doi.org/10.1038/s41598-022-08732-w
SHOW MORE