Abstract

If seismic interferometry is to be applied to a strong ground motion record, the number of available records is limited, compared with the number of ambient noise records. To begin with, we investigated the characteristics of the autocorrelation function for various strong motion records, and the influence of the component of strong motion records to be used for autocorrelation analysis, in order to maximise the number of records available. We found that the autocorrelation function is not influenced by the event or the component of the strong motion record to be used for the analysis, so that reflected waves can be emphasised when a number of autocorrelation function images are prepared without selecting specific types of record. During autocorrelation analysis, we used spectral whitening that worked to remove frequency-dependent source and path effects. In this way, reflected waves from the seismic bedrock could be observed clearly in the whitened autocorrelation function. In addition, we showed the possibility of using the autocorrelation function for exploration by tuning the S-wave velocity structure model to fit the theoretical Green's function to the stacked autocorrelation function. We also found a relationship between the bandpass filter range in the autocorrelation process and the contribution of the multilayer boundary to the reflected wave, from both theoretical and observational examination. We understand the procedure for selecting an appropriate filter range to observe reflected waves from a desired layer boundary.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call