Amplitude Changes of the Seismic Reflected Phases from the Plate Interface after the 2011 Tohoku Earthquake around its Northern Limit Region as Revealed by Active Seismic Surveys

Seismicity along the Japan Trench off the northeastern part of Japan is not uniformly distributed, but it shows spatial variation. Such variation includes a notable seismic-aseismic boundary off Miyagi prefecture around 39°N. Active seismic surveys across this boundary in 1996 and 2001 revealed good anti-correlation between seismicity and amplitude of plate-interface reflected phases such that large amplitude reflections were observed within aseismic regions. By qualitatively reproducing reflection amplitudes by numerical simulations, they concluded that there should exist a thin low P-wave velocity layer over the subducting oceanic plate. They proposed that the existence of abundant fluid is the cause of such low P-wave velocities within the aseismic regions. In 2011, the devastating Tohoku-oki earthquake occurred along the Japan Trench with its fault dimensions reaching 500 km along strike and 200 km along dip. A number of models for its rupture region have been proposed, and all share the common northern limit of the co-seismic slip at around 39°N where considerable contrast of seismicity exists. In order to investigate if physical properties along the plate interface may have changed in response to the rupture propagation, we conducted a seismic survey in 2013 along the same station and profile configuration as that in 2001. We derived 2D P-wave velocity structure along the profiles, and identify reflection phases from the plate interface. We, then, compared their amplitudes between the 2001 and 2013 surveys. Because the type of OBSs are mutually different between the surveys even at the same station sites, and the source signature of the airgun array was also different, we normalized amplitudes of plate-interface reflected phases by the first-arrival refracted waves. The amplitudes of the plate-interface reflected phases appeared decreased within the aseismic regions where abundant fluids had been proposed to exist along the plate interface, whereas they increased within the seismically active region. Such amplitude changes may have been caused by fluid migration during the Tohoku-oki earthquake.