No abstract

No abstract

.[1] We performed shear wave splitting analyses and calculated v P /v S -ratios on local earthquakes recorded at seven stations around Aso volcano, Japan, between 2001 and 2008. The mean v P /v S of 1.61 ± 0.01 from all stations is among the lowest average values found anywhere in the world. A temporal change of shear wave splitting fast polarization is observed at one station. Additionally, spatial variations of shear wave splitting delay times and v P /v S -ratios show a correlation at three stations. We suggest that stress-controlled anisotropy dominates at these stations. GPS baselines and 2D strain reveal extension in the caldera in 2003, and compression before and after 2003. Strain shortening axes indicate either a variation in regional stress or a local stress source that overprints the regional field in 2003. Baseline changes correlate with periods of increased low frequency seismic activity located deeper than 15 km. Compaction and degassing of a previously reported sill at approximately 15 km depth could lead to constantly high pressure of volcanic gases over a wide area and explain the generally low v P /v S -ratios and the observed compression in the caldera. We propose that a small volume of magma intruded into the sill in 2003, caused extension at the surface, and triggered ash eruptions between 2003 and 2005. Associated localized migration of gases in overpressured cracks with aspect ratios potentially ∼0.5 in the upper crust with porosities ≥0.1 could be the source of changes in seismic anisotropy, and explain the slight variations observed in v P /v S .

The first four months of aftershocks of the Darfield earthquake have been studied using data from temporary and permanent seismic stations to investigate the fault geometry, stress field and evolution of seismicity and seismic properties. Earthquake relocations illuminate fault segments and show that the majority of aftershocks occurred beyond the areas of highest slip during the Darfield earthquake. Seismic anisotropy shows a mixture of fast directions parallel to the maximum horizontal stress and fault-parallel fast directions. This, combined with the lack of observable growth of seismicity along fault segments, suggests that the Greendale Fault broke a pre-existing fault plane.