Although there are many studies that deal with complex slip distribution or rupture propagation on an earthquake fault, they usually regard a fault system as a fault of simple geometry. Actual fault systems have highly heterogeneous slip distribution and very complicated shapes, as is often observed through field surveys of surface breaks. In this study, we synthesize seismograms including static displacement near a fault using the discrete wavenumber method in order to estimate the effects of the above types of fault complexity in a quantitative manner. We introduce a complex slip distribution based on the Nojima Fault associated with the 1995 Hyogo-ken Nanbu earthquake. As a result, we show that strong motions at a frequency of lower than 1.0 Hz are strongly affected by the complexity of the fault geometry, at a scale of not more than several km, rather than the rupture propagation style. Distributions of static displacement fluctuate, depending on the fault geometry characterized by the length of each fault segment. Such small-scale variations in fault geometry (≤1 km) have been mostly ignored prior to this work. Our results also suggest that details of fault segmentation and bending can be determined by dense observations (e. g., GPS or geological surveys) of static displacement near a fault system, indicating the importance of simultaneous studies on static and dynamic near-fault motions.

View full abstract

This paper is concerned with the reflection and transmission coefficients of SH-waves at a corrugated interface between two anisotropic heterogeneous elastic solid half spaces. Both the half spaces are taken transversely isotropic and laterally and vertically heterogeneous. The Rayleigh's method of approximation is adopted and expressions for reflection and transmission coefficients are obtained in closed form for the first-order approximation of the corrugation. In Rayleigh's method, expressions in boundary conditions containing the function defining the corrugated boundary are expanded in Fourier series and unknown coefficients in the solutions are determined to any given order of approximation in terms of a small parameter characteristic of the boundary. The analytical expressions of these coefficients show that they depend upon corrugation of the interface and are strongly influenced by the anisotropy and heterogeneity of the half-spaces. Numerical computations are performed for the case of a particular corrugated interface: ζ = c cos k*x showing that the effect of heterogeneity on the reflection and transmission coefficients is minimum near the normal incidence and dominance of this effect increases with the angle of incidence. For incident wave striking at 45°, the effect of the corrugation is found significant on the reflection and transmission coefficients. The maximum effect of transverse isotropy on the reflection and transmission coefficients is observed at normal incidence when the values of the anisotropy parameters are 0.5 and 0.8 for the upper and lower half-spaces, respectively. The effect of frequency of the incident wave is observed on all reflected and refracted waves. The analytical expressions derived by Tomar and Saini (1997), Gupta (1987) and Asano (1960) are obtained as particular cases with our formulation.

View full abstract

A seismic experiment was carried out in east Shikoku, Japan, to detect deep reflections across the Median Tectonic Line (MTL), which juxtaposes low-P/T metamorphic rocks with high-P/T metamorphic rocks. Our experiment employed an unconventional technique: sparse shot spacing, a strong energy source (dynamite) and a dense array of seismometers. The above specifications produce only single fold coverage without common midpoint (CMP) stacking. Nevertheless, the reflection profile provides essential information on the deep structure of the MTL, of other major faults, and of the Moho in east Shikoku. On the MTL, this profile is the first to delineate the MTL from the surface to about 12 km depth. The following three factors were essential to the success of our experiment. First, the receiver interval was sufficiently small to provide horizontal resolution that was able to detect deep reflectors. Second, the simple crustal structure does not require CMP stacking to enhance data quality. Third, a thin weathering layer at the surface reduced the attenuation of seismic waves and minimized the generation of the surface waves that often obscure deep reflectors. In these conditions, the technique can be an effective means of probing the deep crust while substantially reducing survey costs.

View full abstract

During the 1977 eruption of Usu volcano, magma was emplaced at shallow crust. This intrusion induced fumarole activity immediately after the eruption. Based on the repeated thermal observations, the amount of heat discharged by this thermal activity is estimated to be 2 × 10¹⁷J. The corresponding volume of the intrusion is 6 × 10⁷m³. The inferred intrusion volume is comparable to the volume of the resistive block beneath the major faults formed by this eruption, which was interpreted as a cooled intrusion on the basis of recently conducted MT surveys. The heat discharge rate is a surface boundary condition for an underlying magma hydrothermal system. A mathematical simulation, which accounts for multiphase mass and heat transport within a porous media, is conducted to reproduce the thermal activity of Usu volcano. The simulation incorporates the supply of latent heat by solidifying magma and heat transfer by degassing. Permeability conditions are important factors to fit the simulated heat discharge rate with the observation. Increased permeability of surrounding formations causes early appearance and high amplitude of surface heat discharge rate and reduced permeability causes opposite effect. The intrusion permeability has a strong influence on the surface thermal activity. High permeability is needed for the early appearance of the surface heat discharge rate, although it results in the maximum intrusion temperature that is much lower than the observed fumarolic temperature. To avoid the inconsistency, temperature dependent permeability is used in the simulation. The inside of the intrusion with the temperature dependent permeability that has low initial values and the high permeable margin are supposed to satisfy the conditions of the observed surface heat discharge rate and fumarolic temperature.

View full abstract

Left-handed, circular polarized ULF waves with frequency of 0.3-1.1 Hz were detected by GEOTAIL at 27 lunar radii upstream of the moon when the spacecraft was magnetically connected with the lunar wake. The wave was detected twice at 16: 45-17: 00 and 18: 55-19: 02 on October 25, 1994, when the spacecraft and the moon were on the dawn side of the Earth's magnetosphere. The ULF wave was propagating in a direction nearly parallel to the background magnetic field. The observed frequency and polarization are explained by reversal of polarization of right-handed, sunward-propagating electron whistler waves with frequencies above 1.4 Hz in the solar wind frame of reference, which were excited through the interaction with electron beams flowing in anti-sunward direction downstream of the lunar wake. The downstream flow of electron beam is explained by filtering effect of the potential drop at the boundary of the lunar wake. Low-energy components of electrons are reflected back by the potential drop, and the rest components, with energies higher than that of the electric potential penetrate through the wake. The velocity distribution of downstream electrons would be modified to have some bump or shoulder in energy range to form a beam, which is likely to excite whistler mode wave through cyclotron resonance. The lowest energy of the resonant electrons was calculated to be 0.96-2.5 (keV) from the lower boundary of the detected frequency. The variation in the lowest frequency suggests that there are some regions of the lunar wake where potential drop is reduced.

View full abstract