A plate motion model around Japan
We estimated a precise plate motion model for tectonic plates surrounding Japan with the up-to-date GPS velocity fields which are alighned in ITRF 2000. Errors associated with velocity fields are estimated using the empirical relationship between WRMS and white and flicker noise amplitude. The obtained plate motion model generally agrees with those proposed in the previous study. We did not find deformation fields associated with the super plume under French Polynesia, presumably because of the decoupling between the low viscosity mantle and the plate or simply because hot mantle does not reach the above plate. TNGA, which lies on the Australia plate, has a significant differential velocity with the Australia plate. This is due to the rapid spreading of the Tonga-Lau basin. The site velocities at SUWN and DAEJ have significant differential velocities with the Eurasia plate. Therefore, we need to introduce the Amur plate or a regional block to explain these velocity fields. The site velocities at MAG0 and PETP have significant differential velocities with the North American plate. Therefore, we need to introduce the Okhotsk plate or some regional block motion to explain these velocity fields. Site 2003 seems to have significant differential velocities with the Philippine Sea plate. However, this is not conclusive since the GPS time series has a large WRMS.