T12C-02
Vp structure in the largest slip area of the 2011 Tohoku-oki earthquake by airgun-ocean bottom seismometer surveys

Monday, 14 December 2015: 10:35
304 (Moscone South)
Ryosuke Azuma1, Ryota Hino2, Yusaku Ohta2, Kimihiro Mochizuki3, Yoshio Murai4, Yoshihiro Ito5, Hiroshi Yakiwara6, Toshinori Sato7 and Masanao Shinohara3, (1)Tohoku University, International Research Institute of Disaster Science, Sendai, Japan, (2)Tohoku University, Graduate School of Science, Sendai, Japan, (3)University of Tokyo, Bunkyo-ku, Japan, (4)Hokkaido University, Sapporo, Japan, (5)Disaster Prevention Research Institute, Kyoto University, Kyoto, Japan, (6)Nansei-toko Observatory, Kagoshima, Japan, (7)Faculty Science Chiba Univ, Chiba, Japan
Abstract:
The huge coseismic slip during the 2011 Tohoku-oki earthquake concentrated beneath the lower inner slope of the middle Japan Trench (Miyagi region) and its amount sharply diminished along the trench within a hundred kilometers (e.g., Iinuma et al., 2012). The previous 2-D seismic study in the corresponding area found a notable Vp change in the hanging wall side of the megathrust (Azuma et al., in prep.). High-Vp body, corresponding to the rigid backstop block (Tsuru et al., 2002) was identified in the Miyagi, but was absent in the northern area. The spatial extent of the high-Vp anomaly almost coincides with the large slip zone and the correlation suggests that the heterogeneous structure in the overriding plate controlled the rupture process of the M-9 mainshock. To confirm the correlation between the seismic structure and coseismic slip distribution, we conducted a wide-angle survey in 2014 around the previous survey.

In analysis, we operated a traveltime inversion for the first arrivals (Fujie et al., 2000) to obtain a rough model. Then, we made a trial-and-error adjustment of the structure model based on the travel time modeling with a 2D ray tracing (Zelt and Ellis, 1988). In the modeling, we tried to identify the location of the boundary between the backstop block and the low-Vp prism, because we assume the boundary (B/P boundary) to be correlated to the border of the high-slip zone during the 2011 event.

The results showed sharp Vp transition from >3.7 km/s (backstop) to <3.4 km/s (prism) representing the B/P boundary. This feature agrees with the Vp variation or the along-trench distribution of the prism clarified by the previous studies. In the several observed record sections, there are clear shadow zone of the refracted arrivals from the backstop layer. The presence of the shadow indicates that a low Vp layer underlies the backstop, which is consistent with the previous reflection profiling made across the trench system. The low Vp prism is located between the backstop and subducting oceanic crust. Further analysis of the appearance of the shadow zone would constrain the along-fault distribution of the low-Vp material, which we expect to correlate with the spatial pattern of the seismic slip.