Search

We propose a method for capturing the near-surface elastic wave velocity variation using microtremors. The proposed method does not directly estimate the velocity of the medium; instead, it measures the spatial velocity change. We apply the stretching method, which is often used to detect the time variation of shear wave velocity based on the monitoring of seismic ambient noise, to microtremor recordings. Because spatially dense ambient noise records are unavailable, we performed microtremor measurements using multiple linear arrays with multiple sensors along the survey line to collect as many recordings as possible. The synchronised microtremor recordings of each array were used to compute the horizontal-to-vertical spectral ratio (H/V), cross-correlation (CC), autocorrelation (AC), and single-station cross-correlation (SC). The H/V ratio and the waveforms of these correlation functions showed significant spatial variation. The velocity variation was measured quantitatively using the stretching method, which correlates a compressed and stretched reference function with each function on the line. The velocity variation of the vertical and radial components of CC compares well with the variation of the H/V peak, indicating that the radial components of CC retrieved the Rayleigh-wave Green's function. The waveforms of AC and SC also showed spatial variation but the measured value of velocity change was smaller than that for CC. Further investigation is awaited to conclude if this is due to low-quality data or inappropriate measurement conditions. The reliability of the velocity change was confirmed by comparing the results to dispersion curves obtained using the conventional array method. [ABSTRACT FROM AUTHOR]

This paper reports the results of a case study of the application of the microtremor exploration method in western Turkey. We have conducted microtremor surveys at strong motion stations in Aydin and Denizli provinces to derive shallow S-wave velocity profiles. Array records of vertical microtremor motion were acquired at 12 and 15 stations in Aydin and Denizli provinces, respectively. A spatial autocorrelation analysis was applied to the records to estimate Rayleigh wave phase velocities in the frequency range of 3–30 Hz. Then, the phase velocity dispersion at each station was inverted to a one-dimensional S-wave profile down to a maximum depth of about 50 m. The soil models at 5 and 11 stations in Aydin and Denizli provinces, respectively, have a bedrock with an S-wave velocity of 400–600 m/s at depths less than 10 m. The spatial distribution of average S-wave velocities in the top 30 m of the revealed profiles shows low values in the Büyük Menderes Graben and the Denizli Basin, and high values in the surrounding hill or mountain areas. We compared our S-wave profiles with those from previous surface-wave studies. S-wave velocities in shallow parts to a depth of 5–10 m are similar in the two studies. However, S-wave velocities in deep parts are different at some of the stations. [ABSTRACT FROM AUTHOR]