Spatial distribution of shear wave anisotropy in the crust of the southern Hyogo region by borehole observations

Abstract

SUMMARY In order to estimate the detailed orientation and density of cracks around an active fault area, we investigate the spatial distribution of shear wave anisotropy in the crust of the southern Hyogo region with high-quality waveform data recorded by a borehole network operated by the Geological Survey of Japan. We introduce the vertically aligned crack model of Hudson (1980, 1981) to explain its anisotropic structure. At stations more than 10 km away from the earthquake fault zone (i.e. Ikeda and Inagawa), crack orientation is nearly consistent with the direction of the regional maximum stress in this area (WNW–ESE). Cracks are distributed at depths below 5 km with a nearly constant crack density, e, of 0.015 at Ikeda, and below 0 km with e=0.009 at Inagawa. These upper-limit depths of crack distribution are roughly comparable with those of microearthquakes there. In contrast, cracks are aligned nearly parallel to the strike of the earthquake fault (NE–SW) at stations at both ends of the fault system (i.e. Takarazuka and Ikuha), and their intensities of anisotropy are three or four times larger than those at Ikeda and Inagawa. At Takarazuka in particular, cracks are distributed at a depth shallower than 4.5 km with a large crack density, e, of 0.06. This result suggests that major stress release by the faulting of the main shock dominates the present stress field at both ends of the fault system. In the middle of the fault system (i.e. Hirabayashi), cracks are aligned parallel to its regional maximum stress and the intensity of anisotropy is twice as large as those at Ikeda and Inagawa. This result contrasts with that of Tadokoro et al. (1999), who determined the crack orientation at a station close to Hirabayashi to be parallel to the fault direction for events in a period earlier than this study by about a year, suggesting that the stress field in this area may have changed over the timescale of a year. Around Inagawa, the distribution of anisotropy seems to have strong azimuthal variations. This result indicates small-scale (of the order of several kilometres) variations of anisotropic structure in the upper crust, probably due to strong heterogeneities in the fluid and/or constitutive materials.