Assessing the Applicability of Deconvolution of Borehole Records for Determining Near-Surface Shear-Wave Attenuation

The near-surface shear-wave attenuation characteristics of a site are crucial for the accurate prediction of seismic ground motion. However, attenuation is usually either approximated by means of empirical correlations or inferred based on limited laboratory data. An empirical method for the assessment of attenuation, which uses high-quality borehole earthquake data, has been proposed by Fukushima et al. (2016). The method is based on a deconvolution analysis of the seismogram recorded downhole and at the wellhead, which allows the separation of the incident and surface-reflected waves on the deconvolved time series. A frequency-dependent attenuation is then estimated from the transfer function between the incident and reflected waves. In this study, we use 1D synthetic wavefields to perform a sensitivity analysis to (a) test the effectiveness of the method in its original form, as well as with some methodological modifications; (b) examine the effect of different parameters involved in the methodology such as downhole sensor depth, existence of a high shear-wave velocity contrast, and input signal duration on the applicability of the method; and (c) explore the potential extension of the originally suggested application criteria with respect to the minimum depth of the borehole, which essentially originate from the capability of the method to separate the incident and reflected waves on the deconvolved time series. The applicability of the method is characterized by the frequency band for which the method results in satisfactory estimates of attenuation. The method is shown to be quite robust with respect to different parameters regarding both the inherent characteristics of the site under investigation and the available simulation alternatives. Results are promising for extending the application of the method to sites of relatively shallow borehole instrumentation.