FAULT-ZONE STUDIES OF THE M9 TOHOKU-OKI, JAPAN EARTHQUAKE SEQUENCE

California Institute Of Technology, Pasadena CA

Investigators

Abstract

One of the grand challenges for the seismology community involves slippage along plate boundaries. Energy released from many years of locking along subduction zones is made particularly difficult to understand since they occur beneath the deepest portions of the oceans. But they are very important because large subduction zone earthquakes pose one of the greatest natural hazards to mankind, either by strong shaking, flooding (tsunami), or a combination of both, i.e., the recent Tohoku-Oki Earthquake. Rupture properties of these rare events are poorly understood, especially Tsunamigenic Earthquakes (TE) with little pre-tsunami shaking such as the 1896 Sanriku earthquake. The well recorded 2011 Tohoku-Oki earthquake sheds some new light on these issues where a generic interplate event at depth appears to have triggered a rare TE event. Recent reports indicate that the strong shaking came from rupture at depth >30km with relatively small magnitude (Mw~8.5) with 3D amplification occurring primarily at relatively short-periods (2-10 sec) and higher, whereas, the flooding was caused by the large displacement (Mw~9) close to the trench (>50m) and was dominated by relatively long-period (> 20 sec) radiation. The mixture of these rupture styles presents significant challenges in the analysis and modeling of this earthquake and requires data sets covering a broad range of periods. Fortunately, the data coverage on the Japanese Islands is remarkable providing the best set of observations ever recorded for both the main event and the many aftershocks. The investigators will use high-resolution waveform modeling to investigate the variation in fault zone structure in an attempt to understand the great variation in rupture properties both along strike and as a function of depth. The work is a program of technique development along with broadband inversions of the Tohoku-Oki earthquake sequence. The basic aim is to address the locations and source parameters of these events embedded in a 3D structure, exploiting our semi-automated Cut-and-Paste (CAP) method for regional data (0 to 400 km) and CAPt at teleseismic distances. Joint inversions will be used to establish the most transparent paths to arrays for extending the broadband inversions of the main event and large foreshocks and aftershocks.