Surface and Body Wave Extraction from Ambient Noise Data Affected by Strong Vertically Incoming Storm Noise
Abstract
We present the early results of an ongoing project to analyze an ambient noise data set provided by Shell. The aim of this project is to go beyond the well-established practice of inverting for shear velocity structure using surface wave dispersion curves extracted from noise cross-correlations, as well as to maximise usable information in the given data. Specifically, the data present a challenge in that only 10 out of the total 90 days of data produce visually useful looking noise correlation functions after standard processing techniques. The remaining 80 days produce cross-correlations with high amplitudes around zero time lag, suggesting vertically incoming signals from below the receiver array. These low-frequency signals are believed to originate from tropical storms about 90 degrees away. We employ special processing techniques, currently in the test phase, to suppress the effects of these storms and enhance the recovery of propagating waves from the majority of our data. For all days, we exploit the three-component nature of the data to extract Rayleigh wave ellipticities as well as body waves through polarization analysis, with the future aim of constraining Vp and possibly density for the region of study. Azimuthal noise directivity is accounted for by beamforming as well as a less commonly used kernel-based source inversion method.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFM.S13D0493D
- Keywords:
-
- 7299 General or miscellaneous;
- SEISMOLOGY