Variability of Megathrust Earthquakes in the World

A megathrust earthquake of magnitude (Mw) 9.0 hit the Tohoku-Kanto District of Japan, on 11 March 2011. The Pacific coast of the Tohoku district is one of the greatest areas of focus for the earthquake prediction program of Japan. This area's seismicity, in the past and the present, has been investigated in detail and characterized into regional seismic segmentations. The 2011 megathrust earthquake ruptured almost all of the segments between the Pacific coast of the Tohoku district and the Japan trench, causing devastating tsunamis. Contiguous segments of the fault spanning more than 500km broke at once in the earthquake, rather than one or at most two, as had been assumed. How did this event grow to such a scale? The prime factor that had not been recognized before is the double segmentation along the Japan trench (the distinction of shallow and deep segments perpendicular to the trench axis). The apparent absence of earthquakes in the trench-ward segments led us to assume that the plate interface there could be slowly but steadily slipping without building up any strain, whereas the deep segments, the Japan Island-ward segments, failed repeatedly in smaller earthquakes. Slip deficit for more than 500 or 1000 years along the trench-ward segment would have induced the strong initial rupture of the earthquake, which has caused a secondary rupture in the surrounding segments. The subducting plate in this region is one of the oldest plates in the world. This leads us to conclude that the 2011 Tohoku megathrust earthquake belongs to a different class of great earthquakes such as the 1960 Chile, where a young and buoyant plate is subducting rapidly under the continental plate. The earthquake is not similar to the 2004 Sumatra megathrust, which occurred along the oblique subduction zone with the focal mechanism of a low-angle thrust fault. A comparative discussion will be further made for the 1952 Kamchatka, the 1957 Andreanof, the 1960 Chile, the 1964 Alaska, the 1965 Rat Island, the 2004 Sumatra and the 2010 Maule earthquakes, considering down-dip single/double segmentations, seismic quiescence prior to earthquakes (seismic gaps), focal mechanisms, rupture patterns and back-arc activities.