Strain anomaly of the inland crust of NE Japan induced by the 2011 off the Pacific coast of Tohoku Earthquake (Mw9.0), Japan

It is known from a long-term deformation in NE Japan that a strain concentration zone occurs along the Ou-backbone range. The strain concentration zone is though to be an indication of stress transfer to elastic upper-crust from relaxed lower-crust of low viscosity. If this is the case, the strain concentration zone must be detected as an insufficient E-W extension zone from a step-wise strain changes associated with the 2011 off the Pacific coast of Tohoku Earthquake (Mw9.0). We, then, investigated crustal strain changes induced by the earthquake. The earthquake ruptured the plate interface of as long as about 500km, from off Iwate to off Ibaraki. The observed coseismic displacements of GPS stations operated by GEONET, JNES, Mizusawa VLBI Observatory and Tohoku University were as large as up to 5.4 m at the east coast of NE Japan. They were several tens cm even along the west cost being about 400 km away from the source. The observed geodetic signals were thus significant over the NE Japan. Orientations of the displacement vectors were relatively uniform within E-ESE in the central NE Japan, suggesting that the induced stress changes in this area were uniform. We first determined a source model consisting of two rectangular faults with uniform slip to reproduce the observed displacements under assumption of an isotropic homogeneous elastic media. Residual strain changes were, then, obtained by subtracting strain changes calculated from the model from those evaluated from the observed displacements. Because magnitude of the strain changes induced by the earthquake decreases with an increase in distance from the source, amplitude of strain anomaly should be small in the distance, even when a degree of anomaly in rheology of the crust is the same. We, therefore, calculated a ratio of observed strain changes to theoretical strain changes to compensate the spatial variation of induced strain changes. It was found in both of the residual strain changes and the strain change ratios that the E-W extension in the Ou-backbone range was smaller than those expected from the source model in a isotropic homogeneous media. This evidence strongly supports the hypothesis that viscosity of the lower crust beneath the Ou-backbone range is low. On the other hand, we found an overextension zone on the west of the strain concentration. This cannot be explained by such a simple rheology model of the strain concentration zone as stated above. Because we adopted a simple source model consisting of rectangular faults with uniform slip, it is possible that a strain inhomogeniety due to a slip inhomogeniety on the fault is misidentified as the strain anomaly. We, then, evaluated the residual strain changes and the strain change ratios using an inhomogeneous slip model obtained by Iinuma et al. (2011). The strain anomalies of underextension along the strain concentration zone and the overextension on the west were well conserved against the slip inhomogeniety. Therefore, the strain anomalies detected in this study are not artifact due to the simple source model. These findings provide new clues to establish a rheology model in NE Japan.