Shear-wave splitting in the northeastern Japan and southwestern Kurile arcs and its implications for mantle corner flow

This paper investigates shear-wave splitting beneath the southwestern (SW) Kurile arc (Hokkaido) and the northeastern (NE) Japan arc (Tohoku) using local S phases to constrain a spatial pattern of mantle flow generated by the subduction of the Pacific plate. This region includes the arc-arc junction, where the direction of the plate motion is changed relative to the trench axis, and is an excellent natural laboratory for studying the effect of local slab geometry on mantle corner flow. The oblique subduction is taking place beneath the SW Kurile arc in contrast with the normal subduction beneath the NE Japan arc [DeMets, 1990]. It is known that the plate dip angle beneath the land area varies along the arc; 27-30 degrees in NE Japan and 40-50 degrees in the SW Kurile arc [e.g., Umino et al., 1995; Katsumata et al., 2003]. Detailed investigation of shear-wave splitting in subduction zones would provide important information concerning tectonic deformation and dynamic process of the mantle wedge. Waveforms of 1267 intermediate-depth earthquakes recorded at 386 seismic stations were used, and 4729 splitting parameters, the leading shear-wave polarization direction (fast direction) and delay time between two split waves, were observed. Obtained results show that most of fast directions observed at stations in the back-arc side are nearly perpendicular to the trench-axis, whereas those at stations in the fore-arc side are sub-parallel to it. Average delay times observed at stations in the back-arc side are 0.2 - 0.5 s, and those at stations in the fore-arc side are smaller (0.05 - 0.1 s). This systematic spatial variation is observed for the whole study area. We infer that the anisotropy caused by lattice-preferred orientation of olivine, which is probably attributable to the upwelling flow potion of the mechanically-induced convection in the mantle wedge, is a likely candidate for the shear-wave splitting in the back-arc mantle wedge. If this is the case, the present observations may indicate that the upwelling return flow direction is sub-parallel not to the relative plate motion direction but to the maximum dip-direction of the subducting slab beneath the SW Kurile arc.