Direct-path acoustic ranging across the Japan Trench axis, Adjacent to the Large Shallow Thrusting in the 2011 Tohoku-Oki earthquake

Seafloor geodetic data, i.e. GPS/acoustic measurement and continuous seafloor pressure monitoring, brought important evidences showing that the 2011 Tohoku-oki earthquake (Mw 9.0) caused huge (> 50 m) coseismic slip near the Japan Trench. The postseismic behavior of the large slipped area is required to clarify to understand why large amount seismic slip could occur there. We started making direct-path acoustic ranging across the trench axis to reveal the convergence rate between the subducting Pacific and overriding continental plates. We expect the change of the baseline length across the trench axis, the plate boundary, reflects the slip rate at the shallow megathrust, which is difficult to estimate only from other geodetic observations largely affected by intraplate deformation caused by the postseismic viscoelastic relaxation process. 　To this end, we developed an ultra-deep seafloor acoustic ranging system. Our previous ranging systems have been designed to measure baseline length ~ 1 km and to be deployed up to 7,000 m water-depth (Osada et al., 2008, 2012). In order to realize the measurement across the Japan Trench, we improved this system to enhance range of acoustic ranging as well as operational depth of instruments. The improved system was designed to allow acoustic ranging up to 3 km and to be durable under the high-pressure equivalent to water depth of 9,000 m. In May 2013, we carried out a test deployment of the new ranging system. The system is composed of three seafloor instruments equipped with precision transponder (PXPs). Two of the PXPs were set on the landward slope of the Japan Trench, where large coseismic slip happened in 2011. Another PXP was deployed on the seaward side of the trench so that the baseline change associated with the slip on the plate boundary fault, if any, can be detected. Continuous records of baseline lengths were successfully obtained for four months. The repeatability of the distance measurements was about 20 mm for each of the two baselines. Although the duration of the observation was not long enough to estimate precise rate of baseline length changes, it is unlikely that the shortening rates of the baseline lengths exceed the rate of plate convergence (~ 8 cm/a). The results do not support occurrence of evident afterslip along the shallow plate boundary fault in 2013.