2:15 PM - 2:30 PM
[STT38-03] Various field data applications of cSPM analysis for comprehensive evaluation of 3D particle motion
Keywords:3D particle motion, time-frequency analysis, polarized wave detection
The current study advances particle motion analysis within the time-frequency domain through the Spectral Matrix (SPM) analysis, which delineates the 3D particle motion by utilizing eigenvalue decomposition of the SPM. Previously, SPM analysis primarily delineated linear particle motion, typically observed at the arrival of direct P-waves. This study introduces an analysis of time-delay components, culminating in the formation of the complex SPM (cSPM), which addresses the mathematical limitations inherent in previous SPM analyses. By incorporating two time-delay components, the rank of the cSPM is elevated from 2 to 3. This modification facilitates the use of three orthogonal complex bases for the detailed characterization of 3D particle motion, alongside the assessment of in-phase and out-of-phase components. Consequently, cSPM analysis now extends its capabilities to evaluate not only the linear motion of P-waves but also the planar motion of S-waves and other phases, such as converted wave, surface wave and coda waves. We showcase the applications of cSPM analysis across various field data, providing a comprehensive evaluation of particle motion characteristics.
cSPM analysis enhances the characterization of different polarized waves and the identification of low-SNR events. A novel weighting function, derived from the phase information of the first eigenvector, has been introduced to refine the polarization assessment of P-wave arrivals. The efficacy of this approach is evidenced through tests on synthetic waveforms, which reveal an improved distinction between signal and noise, and the detection of two low-SNR events at the Groningen gas field in the Netherlands, previously unnoticed by traditional methods. This strategy proves to be resilient against noise, demonstrating its potential in identifying coherent signals at low SNR levels and in the efficient characterization of polarized waves.
Moreover, cSPM analysis evaluates the planarity and perpendicularity of S-wave polarization relative to the direction of P-wave polarization. Integrating these insights, our methodology not only detected S-wave arrivals in low SNR events but also ascertained the P-S travel times. By applying this technique to four hours of field data from the Groningen field, we successfully identified P-S travel times for cataloged events and additional, previously undetected occurrences, thus facilitating the precise localization of hypocenters via the Rapid Earthquake Association and Location (REAL) method. This cSPM-based detection and location analysis was further extended to two months of continuous data from the same field.
The cSPM analysis was also applied to entire seismograms of microseismic and intermediate-depth natural earthquakes. This enabled the identification of different phase arrivals (P-, PS-converted, and S-waves) by leveraging the magnitude, phase information, and orientation of the first eigenvector (v1). The particle motion shape evaluation through cSPM analysis underscores its wide applicability to various phase characterization challenges. By integrating cSPM analysis information, one can design characteristic functions tailored for diverse detection purposes. Notably, cSPM analysis retains the phase information among 3C waveforms, highlighting the elegance of this newly developed technique.
cSPM analysis enhances the characterization of different polarized waves and the identification of low-SNR events. A novel weighting function, derived from the phase information of the first eigenvector, has been introduced to refine the polarization assessment of P-wave arrivals. The efficacy of this approach is evidenced through tests on synthetic waveforms, which reveal an improved distinction between signal and noise, and the detection of two low-SNR events at the Groningen gas field in the Netherlands, previously unnoticed by traditional methods. This strategy proves to be resilient against noise, demonstrating its potential in identifying coherent signals at low SNR levels and in the efficient characterization of polarized waves.
Moreover, cSPM analysis evaluates the planarity and perpendicularity of S-wave polarization relative to the direction of P-wave polarization. Integrating these insights, our methodology not only detected S-wave arrivals in low SNR events but also ascertained the P-S travel times. By applying this technique to four hours of field data from the Groningen field, we successfully identified P-S travel times for cataloged events and additional, previously undetected occurrences, thus facilitating the precise localization of hypocenters via the Rapid Earthquake Association and Location (REAL) method. This cSPM-based detection and location analysis was further extended to two months of continuous data from the same field.
The cSPM analysis was also applied to entire seismograms of microseismic and intermediate-depth natural earthquakes. This enabled the identification of different phase arrivals (P-, PS-converted, and S-waves) by leveraging the magnitude, phase information, and orientation of the first eigenvector (v1). The particle motion shape evaluation through cSPM analysis underscores its wide applicability to various phase characterization challenges. By integrating cSPM analysis information, one can design characteristic functions tailored for diverse detection purposes. Notably, cSPM analysis retains the phase information among 3C waveforms, highlighting the elegance of this newly developed technique.