The Ministry of Education, Culture, Sports, Science and Technology has employed the microtremor survey method (MSM) for conducting the project aimed at determining the subsurface structures in a sedimentary basin for strong ground motion modeling. The project planned as a national research program after the 1995 Kobe earthquake has been put in practice from 1998 to 2004 in the eight examined areas in Hokkaido and Honshu.
 This paper reviewed the project report laying stress on the role of the MSM played in determining the subsurface structures, as well as on the effectiveness of the MSM to estimating the subsurface structures effectual for strong ground motion modeling. Through the review, the paper also summarized the present situation of the MSM and some subjects related to its practical application to the subsurface structure estimation.
 In addition, taking a serious view of the fact that the unusually high phase velocities of waves have been estimated at a site in Tokyo for frequencies lower than 0.2-0.3 Hz as compared with those to be usually expected, the theory of the present SPAC method was reanalyzed returning to its basics on the assumption that the unexpected phase velocities estimated are due to the wave propagation in which the multimode Rayleigh-wave energy is contaminated.

View full abstract

 We applied the CCA method, a technique of analysis recently developed by ourselves, to a set of microtremor seismograms from an array of vertical-motion sensors, installed around a miniature circle of radius 0.3 m. Rayleigh-wave phase velocities were identified with reasonable accuracy up to wavelengths of several ten meters or, in some cases, well in excess of 100 m. This suggests that, when combined with Konno and Kataoka's (2000) approximate method to associate Rayleigh-wave phase velocities with average shear velocities of the shallow soil, the use of miniature circular arrays provides the opportunity to make a quick appraisal of surface soil properties by way of simple microtremor measurements. This approach presents a groundbreaking advantage over existing microtremor methods, in the sense that a tiny patch of ground, well within the arms reach of a single observer, is sufficient for the estimation of average shear velocities down to depths of several ten meters. We applied the proposed approach to array seismograms of microtremors obtained at three different sites in Tsukuba City, Ibaraki Prefecture. The average shear velocities to depths of 10, 20 and 30 m, inferred with the proposed method, were indeed consistent with available PS logs and known geologic structures.

View full abstract

 Microtremor array observations were conducted in the Matsumoto basin, Japan, to know S-wave structures of shallow and deep sedimentary layers down to the basement having an S-wave velocity of about 3km/s. We first estimated 1D S-wave profiles from inversions of Rayleigh wave phase velocities that were derived from array data of vertical microtremors at 8 sites. Then we performed a long-term continuous observation of three-components of microtremors at two of the 8 sites in the area. The data were processed with a seismic interferometric procedure to retrieve the Green's function between the two sites. The group velocity of Rayleigh wave was estimated from the vertical component of the Green's function. The group velocity agrees with the calculated one for fundamental Rayleigh waves in the 1D profile from the microtremor array exploration. We also estimated group velocities of surface waves from the radial and transverse components of the Green's functions. They are similar to calculated group velocities for fundamental Rayleigh and Love waves for the inverted model. The ellipiticity of observed Rayleigh waves also agrees with theoretical one.

View full abstract

 A surface-wave method and micro-tremor array measurements have been carried out at Central Anamizu, Ishikawa Prefecture, Japan in order to investigate local site effect of earthquake motion in terms of near surface S-wave velocity model. Many houses were collapsed or severely damaged by the Noto Hanto Earthquake in 2007 at the central Anamizu. Collapsed or damaged houses were concentrated around the K-NET Anamizu strong ground motion observation station. It seems that the near surface heterogeneity may strongly affect surface ground motion during the earthquake. The investigation results show that S-wave velocity down to the depth of 20m has strong heterogeneity at the site. A low velocity layer, in which S-wave velocity is 60m/s, beneath the K-NET disappears at the other side of the park only 50m away from the K-NET. Swedish ram soundings were also carried out around the K-NET and the results agreed with one of the surface-wave method and the micro-tremor array measurements very well. Two-dimensional site amplification was calculated by a finite-difference method using the S-wave velocity model obtained through the investigations. The calculation shows that near-surface heterogeneity strongly affects amplification of the site.

View full abstract

 We propose the S-wave velocity structure model through the array analyses of long-period microtremors and the earthquake ground motions retrieved by strong-motion instruments in the southern area of Shizuoka Prefecture, vicinity of Omaezaki and Makinohara (Sagara) Cities, where long-period ground motions have sometimes been observed from moderately large and shallow earthquakes. This region would be suffered from the anticipated Tokai earthquake; therefore, the ground motion modeling is inevitable for assessing seismic safety in this area. The major concern of the present research is the lack of shallow to deep geotechnial data including seismic bedrock (Vs >3.0 km/s). We performed large array measurement of microtremors in this region and analyzed the data by the SPAC method to obtain the S-wave velocity structure as a first step. Next, we verified the structure model by analyzing surface waves included in the strong-motion records.
 The phase velocity dispersion based on the estimated structure model by array analyses of microtremors correlate well with the observed phase velocity dispersion of surface waves contained in the strong motion records. In addition, the particle orbits of the later arrivals in the strong-motion records match well with the theoretical ellipticity of Rayleigh waves based on the structure model. To confirm the comprehensive validity of the structure model, we carried out simulations of waveforms from the 2000 Miyake-jima swarm earthquakes using discrete wave-number method.
The target area in this study is a part of thick sediments, so called accretionary prism, that have been formed associated with the subduction of Philippine Sea plate. The S-wave velocities of the thick sediments vary from nearly 1 km/s to 2 km/s and the thickness is about 4~5 km. The bedrock S-wave velocity is estimated to be 4 km/s that is significantly higher than that of average upper crust in land. The area of such high velocity contrast potentially dominates long-period ground motion consisted of surface waves.

View full abstract

 The microtremors was observed for the earthquake observation site on the ground from a hard ground with thin surface layer to a thick sedimentary basin aiming to clarify whether to approximate microtremor H/V spectrum ratio to the earthquake amplification characteristic in the surface layer, and it compared with the earthquake observation data. As a result, both the horizontal and the vertical response characteristics of the surface layer were reflected in microtremor H/V spectrum ratios, and it was clarified that the idea of approximating microtremor H/V to the earthquake amplification characteristic in the surface invited a possibility of considerably undervaluing an actual S wave earthquake amplification characteristic. Therefore, it newly proposed the technique for simply presuming the earthquake amplification characteristic of the subsurface layer assuming two layer structural model by paying attention to the predominant frequency and the peak amplitude of microtremor H/V, and using the relation between a predominant frequency and bedrock depth, and the relation between the peak amplitude and the horizontal transfer function. The applicability of this simple estimation method was verified according to the earthquake observation data, therefore it is thought that even the ground from the hard ground to a thick sediment plains can presume a rough earthquake amplification characteristic to the safety side by considering about 1.4 times room, and this estimation method is useful for the presumption result as an outline investigation method that precedes a detailed investigation.

View full abstract

 In order to reduce the earthquake disaster, it is very important to know the strong ground motion after a big earthquake as soon as possible. In Japan, high-density seismograph networks were installed after the Hanshin-Awaji earthquake. The typical spacing between each seismic station is approximately 20 km. However, there are many cities and towns that have only one seismic station. The Hanshin-Awaji earthquake has shown us the limits of governmental help and importance of the self-help by citizen. The density of seismic stations in Japan is not still enough for civilian self-disaster prevention. So it is important to know strong ground motions with higher density.
 In this article, therefore, a new method to estimate high-resolution seismic intensity maps using only two seismic stations data has been developed. The new method is able to estimate seismic intensity of any given point. The new method has been applied to Yokohama City. As the results, high-resolution seismic intensity map has been achieved immediately with good accuracy.

View full abstract

 The earthquake ground motions are distinctly influenced by the surface geology. The acquirement of site effect at the strong motion observation point makes it possible to interpret appropriately characteristics of earthquake records observed at the surface or JMA seismic intensity calculated from the records. In this study, first, microtremor measurements were carried out at the observation points of the seismic intensity information network administered by local governments, the K-NET, and the KiK-net by the National research institute for earth science and disaster prevention, in the Chugoku district. Predominant period of the ground was estimated from spectral ratios of horizontal component to vertical one (H/V spectra) of microtremors, those of earthquake motions and transfer function based on the 1-D propagation theory of SH-waves. The estimation accuracy of each method is verified at the KiK-net sites because the predominant period obtained from the ratios of earthquake motion spectra observed on the surface and in the borehole was assumed to be correct. The estimation accuracy of predominant period of the ground by H/V spectra of microtremors is not so different from that by H/V spectra of earthquake motions. The estimation error amounts to 16-18 %. The accuracy of transfer function based on the 1-D propagation theory of SH-waves is lower than those by the two kinds of H/V methods. The predominant period of the ground at each observation points was determined by using the results of the accuracy verification. The predominant period at most inland sites in the Chugoku district is shorter than 0.3 sec. Several sites on the sedimentary basins on the coast have predominant periods longer than 0.5 sec. Many sites in comparatively populous urban region have predominant periods longer than 1 sec. In addition, relation between the estimated predominant period of the ground and the average shear-wave velocity down to 30 m depth (AVS30 ) was considered. Both have comparatively good correlation. The relation of both was formulated, and the AVS30 was estimated from the predominant period of the ground. As a result, the estimation accuracy was higher than that by the digital land information.

View full abstract