Poster #214, Seismology

A tale of urban seismology: ambient seismic noise, machine learning methods, and seismic hazard analysis at the Seattle basin edge

Natasha Toghramadjian, Laura A. Ermert, & Marine A. Denolle
Poster Image: 

Poster Presentation

2021 SCEC Annual Meeting, Poster #214, SCEC Contribution #11436 VIEW PDF
Sedimentary basins amplify and prolong strong shaking during large earthquakes. The city of Seattle sits atop the Seattle basin, and faces high seismic hazard posed by both the regional-scale Cascadia subduction zone and crustal faults. The Seattle fault zone forms the southern border of the basin, presenting an actively deforming basin edge with a complex internal geometry and velocity structure. Resolving the location and character of this active, urban fault is necessary for rigorous seismic hazard assessment in the city of Seattle. We deployed 100 nodal seismometers across the Seattle fault zone and basin edge in four north-south transects for one month in July 2019. All instruments were... hosted by Seattle residents, and installation included educational involvement with residents and students. We process and perform auto-correlation and cross-correlation analysis of the continuous ambient seismic noise data. Correlations are contaminated by anthropogenic activities, particularly highways, which causes spurious arrivals. To select the optimal correlations to generate the stack, we perform Gaussian mixture model clustering. We automate the clustering of correlations that we attribute to daytime and quiet times (nighttime and weekends). We further investigate if the re-constructed single-station correlations enable us to observe subsurface geological interfaces. In particular, we attempt to characterize the structure of the Seattle fault zone.