Shear-wave splitting observed above small earthquakes in a geothermal area of Japan

Abstract

SUMMARYThe phenomenon of shear-wave splitting has been observed at two seismic stations with three-component seismographs above shallow microearthquakes which occurred at depths of 1-2 km in the Takinoue geothermal area of northern Honshu. An almost uniform alignment of the polarizations of the faster-split shear wave in the direction of ENE-WSW may result from the presence of non-randomly distributed liquid-filled microcracks. Two models of crack distribution reproduce the observed polarization pattern. In the first model vertical microcracks are aligned in the ENE-WSW direction, while the second supposes a distribution of microcracks striking in the ENE-WSW direction but dipping randomly.The ENE-WSW strike of the cracks in both models is consistent with the direction of the ENE-WSW maximum horizontal compressive stress inferred from the analysis of earthquake mechanisms in the study area. It differs slightly from the strike of vertical fractures (WNW-ESE) obtained by hydraulic fracturing experiments at sites about 10 km west of the study area. The tectonic stress regime obtained from an analysis of focal mechanisms of microearthquakes occurring in this area seems to be consistent with the latter model of the crack distribution from the shear-wave analysis, at least at depths less than about 700 m.An analysis of the delay times between the faster-and the slower-split shear waves has revealed a region with anomalously large delay times normalized to ray path lengths. A 3-D P-wave velocity inversion indicates that this region is a low P-wave velocity zone. The low P-wave velocities and high shear-wave birefringence is consistent with a relatively dense distribution of liquid-filled microcracks.Key words: crack-induced anisotropy, geothermal area, Japan, shear-wave splitting I N T R O D U C T I O NA weak distribution of thin circular cracks aligned in a single direction makes the elastic medium containing the cracks have an effective anisotropy of hexagonal symmetry with the symmetry axis parallel to the crack normal (Crampin 1978;Hudson 1980Hudson , 1981. Shear waves propagating through the medium split into two orthogonally polarized shear waves, one polarized in the crack plane, and the other in the plane of the seismic rays' propagation (Crampin 1978). If the cracks are vertical, whether they are dry or filled with watery liquid, the polarization directions of the faster-split shear waves are parallel to the strike of the cracks for seismic rays with small incident angles (a little under 40" for dry cracks, and 30" for liquid-filled cracks) (Crampin & McGonigle 1981). Locus of ground motion in the horizontal plane due to the faster-split shear wave abruptly changes its form as the slower shear wave with orthogonal polarizarization arrives. The locus varies from linear to circular motion when the differential travel time between the two shear waves is small, and has a cruciform pattern for the large differential travel time (Crampin 1978). The parallel alignment of the faster shear-wave polarizations and the ch...

Initial shear wave particle motions and stress constraints at the Anza Seismic Network

“…A leading explanation for these observations has been vertical cracks which strike preferentially in the direction of greatest compressive stress (e.g., Crampin 1987;Shepherd 1990). Many of these studies have noted a general correspondence between the preferred initial shear wave particle motion direction, S, and the estimated regional maximum principal compressive stress direction, 8, (e.g., Kaneshima, Ando & Crampin 1987;Kaneshima, Ito & Sugihara 1988, 1989Savage, Peppin & Vetter 1990;. If this relationship between s^ and 8, holds in all circumstances, such observations would provide useful stress constraints wherever the earth could be sampled with high-frequency shear waves.…”

An analysis of split shear waves observed above crustal and uppermost mantle earthquakes beneath Shikoku, Japan: Implications in effective depth extent of seismic anisotropy

“…Recently, many researchers have proposed the presence of seismic anisotropy in the upper crust [e.g., Crampin et al, 1984;Booth et al, 1985;Buchbinder, 1985;Kaneshima et al, 1987Kaneshima et al, , 1988aKaneshima et al, , 1989Peacock et al, 1988]. Most of them have analyzed three-component seismograms from crustal earthquakes with various azimuths and incident angles, finding out parallel or subparallel alignments of the particle motion upon the leading shear wave arrivals.…”

“…This means that it is impossible to resolve an anisotropic structure in the vertical direction, without analyses of shear waves coming from various focal depths and containing a wide range of dominant frequency. Kaneshima and Ando [1986] and Kaneshima et al [1989] have proposed that the seismic anisotropy which causes the observed shear wave polarizations may not extend down to 20 km. Many earthquakes have taken place from 5 km to 50 km beneath the Shikoku area.…”

“…Many earthquakes have taken place from 5 km to 50 km beneath the Shikoku area. We shall offer another constraint on the effective depth extent of the anisotropy through the shear wave analysis of waveform data with much higher quality than those of Kaneshima and Ando [1986] and Kaneshima et al [1989].…”

Shear-wave anisotropy across the geothermal field of Milos, Aegean volcanic arc

“…With the exception of the Takinoue geothermal field in Japan (Kaneshima et al 1988) the regions studied (e.g. Booth et al 1985;Peacock et al 1988) can be considered to be more homogeneous in velocity or stress structure than in the volcanic geothermal field of Milos.…”