Source Process of the 2011 Tohoku Earthquake Estimated from the Joint Inversion of Teleseismic Body Waves and Geodetic Data Including Seafloor Observation Data: Source Model with Enhanced Reliability by Using Objectively Determined Inversion Settings

The rupture process of the 2011 Tohoku earthquake is estimated from the joint inversion of teleseismic body waves and geodetic data using multiple time-window analysis. To enhance the reliability of the inversion solution, two important inversion-setting parameters that control the spatiotemporal slip distribution, the maximum slip duration of each subfault (T-sd) and the propagation velocity of the first time window (V-ftw), are determined by objective criteria. By performing a semblance analysis of small-aperture seismic array data of near-source strong motion waveforms, T-sd is determined to be 100 s. V-ftw is determined to be 2: 0 km/s from the matching of the teleseismic and geodetic slip models. The spatial resolution of geodetic inversion is significantly improved by using seafloor crustal deformation data in the source area in addition to the terrestrial Global Positioning System network data. The estimated seismic moment and the maximum slip are 3.4 x 10(22) N.m (M-w = 9.0) and 43 m, respectively. The total rupture duration is similar to 150 s. Large slip is seen in the shallow part of the fault surface, including the hypocenter, and its spatial extent along the trench axis is similar to 300 km. This model is consistent with the results of bathymetric surveys and the slip models from tsunami data. The relationships among the fault parameters of the characterized source model of the 2011 Tohoku earthquake suggest the possibility that saturation of fault width occurs for this huge earthquake.