Estimating S-wave attenuation in sediments by deconvolution analysis of KiK-net borehole seismograms

Research output: Contribution to journalArticlepeer-review

9 Citations (Scopus)

Abstract

S-wave attenuation characteristics of sediments are important for accurate prediction of ground motion. High-quality borehole seismic observation data have been accumulated in Japan since the construction of KiK-net, a nationwide network, in the early 2000s. We estimated S-wave attenuation (Q−1 S) values in sediments by applying a new deconvolution method to KiK-net data. First, incident and surface-reflected waves from local earthquakes are separated by deconvolving the seismogram from the bottom of the borehole with the seismogram from the ground surface. Then, Q−1 S values of sediments are estimated from transfer functions (or system functions) between incident and surface-reflected waves that are stacked with respect to available earthquakes. Applying the deconvolution method to the records from KiK-net stations having boreholes deeper than 300 m, we obtained stable estimates of Q−1 S values in the 0.5–10 Hz range at 16 stations. The Q−1 S values decrease with frequency up to about 2–3 Hz but become nearly constant at higher frequencies. By fitting a power-law function to the Q−1 S values at frequencies lower than 2 Hz, Q−1 S values at 1 Hz ranged from 0.02 to 0.2, and the exponent of the power law ranged from −0:12 to −1:00. The obtained Q−1 S values show clear depth dependence. Our results are consistent with Q−1 S values estimated by previous studies, which supports the usefulness of our new method.

Original languageEnglish
Pages (from-to)552-559
Number of pages8
JournalBulletin of the Seismological Society of America
Volume106
Issue number2
DOIs
Publication statusPublished - 2016 Apr

ASJC Scopus subject areas

  • Geophysics
  • Geochemistry and Petrology

Fingerprint Dive into the research topics of 'Estimating S-wave attenuation in sediments by deconvolution analysis of KiK-net borehole seismograms'. Together they form a unique fingerprint.

Cite this