A Transdimensional Framework for Array-based Seismic Phase Detection and Characterization
Description:
Identifying seismic phases present in seismic waveforms and measuring their attributes is essential to seismic modeling and imaging of the planetary interior. For example, robust measurement of ray parameters can differentiate depth phases from surface-reflected ones, yielding more accurate velocity models and temperature and compositional inferences based on those models. In addition, polarization parameters can facilitate phase identification and provide information on wave type, propagation path anisotropy, and subsurface structural variations such as dipping layers or local heterogeneities. In this study, we present a novel transdimensional inversion framework designed to infer arrival time, slowness, and polarization parameters of seismic phases from three-component seismograms collected in either a source array or receiver array setting. Our proposed method leverages a probabilistic Bayesian approach to infer the number of phases that can be robustly identified in noisy seismograms while providing robust uncertainty quantification on their polarization and ray parameter attributes. Synthetic tests show promising results in accurately recovering phase attributes, while preliminary experiments with terrestrial seismic data successfully recover seismic arrivals consistent with traditional vespagram techniques, but without the need for specifying detection thresholds for phase arrivals. This work can be applied to the analysis of Martian seismic data, where single-station observations and limited marsquake coverage present unique challenges to seismic waveform interpretation. Applying this method to marsquake waveforms from the InSight mission could enhance the identification of Martian crustal and mantle phases, providing new insights into the planet's seismic structure and thermal evolution.
Session: Exploring Planetary Interiors and Seismology: Observations, Models, Experiments and Future Missions [Poster]
Type: Poster
Date: 4/17/2025
Presentation Time: 08:00 AM (local time)
Presenting Author: Ziqi
Student Presenter: No
Invited Presentation:
Poster Number: 106
Authors
Ziqi Zhang Presenting Author Corresponding Author zzq@umd.edu University of Maryland, College Park |
Vedran Lekic ved@umd.edu University of Maryland, College Park |
|
|
|
|
|
|
|
A Transdimensional Framework for Array-based Seismic Phase Detection and Characterization
Session
Exploring Planetary Interiors and Seismology: Observations, Models, Experiments and Future Missions