Abstract

Summary. Subcrustal earthquakes occur in a slab-like zone beneath SW Japan in association with the subduction of the Philippine Sea plate. They generate a distinct pair of P and S later phases with anomalously low apparent velocities of about 7 and 4 km s−1, which cannot be explained by a laterally homogeneous velocity structure. These later phases are observed for particular source–receiver geometries: the sources at depths shallower than 50–60 km and the receivers beneath which the subcrustal seismic zone comes in contact with the bottom of the continental crust. We examined the nature of these later phases using two methods. One is to determine two types of apparent velocities, an ‘event-common’ apparent velocity and a ‘station-common’ apparent velocity, which may be different from one another for a laterally inhomogeneous velocity structure. The other is to trace seismic rays numerically for a laterally heterogeneous velocity model. An extensive analysis using these two methods leads to a conclusion that the slab-like seismic zone constitutes a low-velocity channel in the uppermost mantle with P and S velocities comparable to those of the lower continental crust beneath SW Japan. The waves trapped within the low-velocity channel escape from it through its contact with the continental Moho and are observed as a distinct pair of later phases. The head waves guided by the bottom boundary of the low-velocity channel are observed as weak arrivals of initial phases. The thickness (<10 km) and the velocities (VP=7 km s−1, VS=4 km s−1) of the subcrustal seismic zone indicates that it is made up of the ‘untransformed basaltic oceanic crust’ subducted with the underlying lithospheric mantle and that the transformation of basalt to eclogite does not take place in the subducting slab at least down to depths of 50–60 km. We present laterally heterogeneous structural models of the crust and the uppermost mantle along two profiles in SW Japan. The models are characterized by an apparent thickening of the lower continental crust in particular regions due to a direct contact of the subducting oceanic crust to the continental Moho. Such features are qualitatively consistent with the results of explosion seismology and with the Bouguer anomaly pattern in the relevant regions. The oceanic lithosphere with the untransformed ‘light’ oceanic crust at its top might still be buoyant after subduction, a factor which must be taken into account in considering the driving mechanisms of plate tectonics, the slab geometry and the earthquake-related tectonic process.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call