**Generation Mechanism of Giant Earthquakes in Subduction Zones with Smaller-Size Interplate Earthquakes During Interseismic Period**

Takane Hori1, Mamoru Hyodo1 and Shin'ichi Miyazaki2

<sup>1</sup>*Japan Agency for Marine-Earth Science and Technology* <sup>2</sup>*Kyoto University Japan*

#### **1. Introduction**

84 Earthquake Research and Analysis – Seismology, Seismotectonic and Earthquake Geology

Wallace R.E. Notes on stream channels offset by the San Andreas fault, southern Coast

Wells D.L., Coppersmith K.J. Empirical relationships among magnitude, rupture length,

Stanford Univ. Publ., Geol. Sci. 1968. V. 11. P. 6-20.

Р. 9741002.

Ranges, California // Proc. of Conf. on Geol. Probl. of San Andreas Fault System.

rupture area, and surface displacement // Bull. Seismol. Soc. Am. 1994. – V. 84.

In the subduction zone along the Japan trench, northeast Japan, only M 7∼8 earthquakes have occurred in the past hundred years (Yamanaka & Kikuchi, 2004). Furthermore, in the surrounding area of such earthquakes, small repeating earthquakes have occurred (Uchida & Matsuzawa, 2011). Because the occurrence of small repeating earthquakes indicates that the plate boundary is creeping (Uchida et al., 2009), it has been considered that only re-rupture of M 7∼8 slip areas (asperities) surrounded by aseismically sliding region can occur in this subduction zone. However, in 2011, a giant earthquake of magnitude (M) 9.0 (we call this as the 2011 Tohoku earthquake hereafter) occurred in the subduction zone and caused destructive tsunami along the pacific coast of Japanese island (Earthquake Research Committee, 2011). The source area extends more than 500km in trench parallel direction and more than 200km in subducting direction including the past M 7∼8 asperities. It should be noted that the occurrence of the M9 earthquake cannot be explained by the combined rupture of the M=7∼8 asperities. The slip amount in the M9 earthquake is one order larger than that in each M=7 ∼8 earthquake. For example, off Miyagi, the central part of the 2011 Tohoku earthquake, the slip amount was 1.8 m for the 1978 off Miyagi earthquake (Yamanaka & Kikuchi, 2004) and more than 10m for the 2011 Tohoku earthquake [e.g., (Iinuma et al., 2011)]. Hence, one of the key questions provoked by this event is: How could an M9 earthquake occur in a subduction zone in which only M = 7∼8 earthquakes have occurred repeatedly in the past 100 years?

This question, however, arises not only for the 2011 Tohoku earthquake, but also for most of other M∼9 earthquakes. In the 2004 Sumatra-Andaman earthquake, for example, an Mw=9.3 event occurred where M=7 ∼ 8 events had occurred separately in space (M=7.7 in 1941, M=7.9 in 1881, and M ≤ 7.5 in 1881) (Subarya et al., 2006). In southern Chile, the 1575 and 1960 earthquakes were M ≥ 9 giant earthquakes, whose rupture area extended about 1000 km, though two other events in 1737 and 1837 were significantly smaller, with rupture areas limited to about 500 km (Cisternas et al., 2005). Before the 1964 Alaska earthquake, M∼8 earthquakes occurred in 1854, 1855 and 1900 within one of the asperity near Kodiak island (Christensen & Beck, 1994). Such smaller earthquakes occurred not only before M9 earthquakes but also after several decades as follows. For the 1952 Kamchatka earthquake (Mw=8.8∼9.0), M∼7 earthquakes occurred in 1904 also in 1993 within the source area of the 1952 earthquake (Johnson & Satake, 1999). After 29 years of the 1957 Alutian earthquake (Mw=8.6), M=7.7 earthquake occurred within the source area (Johnson et al., 1994). Furethermore, for the 1906 Colombia-Ecuador earthquake (Mw=8.8), three Mw=7.7∼7.9 earthquakes occurred in 1942, 1958 and 1979 (Kanamori & McNally, 1982).

Among the above M9 earthquakes, it is known that some of them have occurred repeatedly as in the southern Chile. For example, in the Kamchatka an M9 earthquake also occurred in 1737 (Johnson & Satake, 1999). For the 2011 Tohoku earthquake, identical tsunami deposite distribution has been found in Sendai area, northeast Japan (Minoura et al., 2001). The estimated recurrence time interval is several hundreds to a thousand years. Hence, we assume that the M9 earthquake occurrence is the fundamental rupture mode in each subduction zone. The key question then becomes: How can M=7 ∼8 earthquakes occur within the source area of an M9 event during the long-term seismic cycle of M9 earthquakes? Based on the concept of a hierarchical asperity model that was applied to the M3 sequence within an M5 asperity off Kamaishi along the Japan trench (Hori and Miyazaki, 2010), we speculate that such events could be explained.

In this chapter, first we will introduce the hierarchical asperity model for M9 earthquakes and explain how to represent them as numerical models. Then, we will describe the simulation results, part of which is identical to our short paper (Hori & Miyazaki, 2011). In discussion, we will introduce the fault strength that is introduced by Nakatani (Nakatani, 2001), and discuss the generation mechanism of M9 recurrence with M=7∼8 earthquakes during the M9 interseismic period. Finally, we will compare our results with observation data, especially for the 2011 Tohoku earthquake.
