Historical seismicity and dynamic rupture process of the 2011 Tohoku-Oki earthquake
Highlights
► Historical seismicity data are used to model plate interface heterogeneity. ► Simple model explains complex rupture processes during the Tohoku-Oki earthquake. ► Cascading ruptures of smaller patches work as a dynamic nucleation process.
Introduction
Earthquake rupture processes can often appear to be very complex. An excellent example is the Mw9.0 11 March, 2011 Tohoku-Oki earthquake, one of the best-recorded earthquakes to date. The spatio-temporal rupture process during this earthquake have been studied by a number of research groups (e.g., Ide et al., 2011, Koketsu et al., 2011, Shao et al., 2011, Suzuki et al., 2011). Despite the complexity of this event, the basic characteristics of slip models developed during these studies are similar in each case. The event was a typical low-angle thrust slip along the interface between the Pacific and the overriding Okhotsk or North America plates. The total duration of the event was 130–200 s, with a maximum slip of 30–60 m, a source area of about 400 × 200 km, and a corresponding seismic moment of 4–6 × 1022 Nm (Mw9.0–9.1). Four main stages of the rupture process were distinguished: a small initial phase, followed by deep rupture for up to 40 s, extensive shallow rupture at 60 to 70 s, and continuing deep rupture lasting more than 100 s (Ide et al., 2011).
The most important feature of this earthquake is a large shallow slip near the Japan Trench, which caused a huge tsunami. This slip was identified by waveform data and confirmed by ocean bottom observations (Fujiwara et al., 2011, Ito et al., 2011). Previous research sought to quantify the amount of interseismic slip due to ambient plate motion (e.g., Hashimoto et al., 2009, Nishimura et al., 2004), but a definitive determination of the status of the shallowest off-shore plate interface could not be achieved due to the low resolution of geodetic instruments and poor azimuthal coverage. The large slip of more than 30 m that occurred at the shallowest part of the plate interface implies that this section of the interface was locked, even if not perfectly, prior to the mainshock, leading to the accumulation of strain energy. Another unusual feature of this event is the timing of the shallowest part of the large slip, estimated to have occurred between 60 and 70 s after the hypocentral time. Given that the hypocenter is located less than 100 km from the trench, a typical rupture velocity of about 2–3 km/s would lead to the surface rupture at the trench at around 30–50 s, indicating that this event involved either very slow rupture propagation or a delay in the rupture process.
Previous studies have examined the spatio-temporal characteristics of high-frequency sources of the Tohoku-Oki earthquake (e.g., Ide et al., 2011, Meng et al., 2011, Wang and Mori, 2011, Zhang et al., 2011). Two high-frequency pulses were identified in observed seismograms at inland stations in Miyagi prefecture. These pulses were radiated at different times from similar locations in the deeper sections of the source region, at greater depths than the hypocenter. The location of these pulses is also close to the rupture areas of the 1936 and 1978 Miyagi-Oki earthquakes. Prior to the Tohoku-Oki earthquake, the Japanese Government, through the Headquarters for Earthquake Research Promotion, released an official long-term forecast for future Miyagi-Oki earthquakes in which the occurrence probability of a MJ (magnitude determined by Japan Meteorological Agency, JMA) 7.5 earthquake from this source area was determined to be 99% over 30 years. The relationship between the actual rupture process during the Mw9.0 mainshock and this anticipated scenario of a MJ 7.5 event needs to be clarified. During the later stages of the mainshock rupture, several intermittent sources of high-frequency waves were identified in the southern part of the source region and close to the coast line. These ruptures may also be related to historical earthquakes.
Another remarkable feature of the Tohoku-Oki earthquake is significant foreshock activity, which started with a Mw7.3 earthquake two days before the mainshock. Using the increased seismicity during these two days, Ando and Imanishi (2011) and Kato et al. (2012) proposed a slow slip event that was undetectable by seismic and geodetic instruments. The initial part of this mainshock was faint compared with the foreshock, suggesting a relatively small stress drop (Ide et al., 2011). Uchide et al. (2011) also indicated that the waveform of the initial part of the mainshock resembled that of a M5.5 earthquake. Together, these observations suggest that the M9 earthquake started from a small rupture and eventually grew to a catastrophic event.
We have already published our preliminary conceptual model (Aochi and Ide, 2011) to address some of the points outlined above. This model utilizes the multiscale heterogeneous model concept developed by Ide and Aochi (2005) and Aochi and Ide (2009), in which a heterogeneous fault plane is described by circular patches of various sizes, and power law size–frequency statistics. These patches are ruptured according to a slip weakening friction law characterized by slip weakening distance, which in turn is assumed to be proportional to the patch radius. Given this, under homogeneous breakdown strength drop, the fracture energy in these circular patches linearly scales with the radius. Aochi and Ide (2011) demonstrated that the observed delay of the shallow large slip was related to the longer time period needed to initiate a rupture in higher fracture energy area within a large patch.
This paper advances the work of Aochi and Ide (2011), which focused on modeling the rupture growth features using only a few patches, assuming a spatially heterogeneous plate interface without constraints based on real data, and did not discuss the contribution of foreshocks to the rupturing process. In comparison, this study uses historical seismicity to map out heterogeneous fault patches. The next section explains our numerical model that connects this historical seismicity to the fault patches used in our modeling. We then present the results of simulations based on several parameter studies that identify critical features of the rupture process. This modeling and simulation suggests that foreshock-derived strain accumulation was an important stage that led to the M9 earthquake. The final section of the paper discusses the use of historical seismic data for long-term earthquake prediction.
Section snippets
Overview of the multiscale circular patch model
During this study, we employ the hierarchical patch model developed by Ide and Aochi (2005) and Aochi and Ide (2009). In this model, the fault plane (i.e., the plate interface in the case of the Tohoku-Oki earthquake) has heterogeneous frictional properties that can be characterized by a number of circular patches. For every patch, a simple slip weakening law was assumed, in which the shear stress during slipping τ is given as a function of slip Δu,where Δτb, τr, and Dc
Incomplete rupture without the rupture of the largest patch
First we consider the patch distribution simply defined by historical earthquakes in Table 1 and the largest patch as shown in Fig. 2. A rupture is developed as a small-scale cascade, causing ruptures in two surrounding fourth-level patches, although this rupture is arrested by the area of large fracture energy area defined by the largest patch. Fig. 3 shows snapshots of velocity, slip, and stress. The obtained moment rate function is similar to the slip model of Ide et al. (2011), suggesting
Discussion
In the previous sections, we tested only one homogeneous stress condition, in which the breakdown strength drop and dynamic stress drop are 10 and 4 MPa, respectively. Although the stress drop is mainly constrained by the size of fault and seismic moment, we have larger ambiguity on the value of strength drop. The value 10 MPa was selected to reproduce the characteristics of the Tohoku-Oki earthquake. Fig. 7 summarizes the result as moment rate functions using other selections for this parameter.
Conclusions
This study considers a hierarchical structure as an essential property of rupture phenomena, including earthquake source dynamics. A set of discrete circular patches and the dynamic rupture model proposed by Ide and Aochi (2005) can successfully explain several features of the rupture process of the Tohoku-Oki earthquake. These include an initial phase of small-scale irregular rupture, followed by initial downward rupture propagation until 40 s, a delayed main rupture near the trench, a second
Acknowledgments
Computer systems at the Earthquake Information Center of the Earthquake Research Institute, the University of Tokyo and the French National Computing Center GENCI-CINES (grant 2012-46700) were used during this study. Figures were prepared using GMT software (Wessel and Smith, 1998). This work was supported by JSPS KAKENHI (23244090), MEXT KAKENHI (21107007), and French–Japanese ANR-JST project DYNTOHOKU (2011–2013).
References (42)
- et al.
A unified source model for the 2011 Tohoku earthquake
Earth and Planetary Science Letters
(2011) - et al.
NIED seismic moment tensor catalogue for regional earthquakes around Japan: quality test and application
Tectonophysics
(2002) Dynamic stress drop and rupture dynamics of the October 15, 1979 Imperial-Valley, California, earthquake
Tectonophysics
(1990)Maximum likelihood estimation of b in the formula log N = a − bM and its confidence limits
Bulletin of the Earthquake Research Institute
(1965)- et al.
Possibility of Mw 9.0 mainshock triggered by diffusional propagation of after-slip from Mw 7.3 foreshock
Earth Planets Space
(2011) - et al.
Three-dimensional nonplanar simulation of the 1992 Landers earthquake
Journal of Geophysical Research
(2002) - et al.
Numerical study on multi-scaling earthquake rupture
Geophysical Research Letters
(2004) - et al.
Complexity in earthquake sequences controlled by multiscale heterogeneity in fault fracture energy
Journal of Geophysical Research
(2009) - et al.
Conceptual multi-scale dynamic rupture model for the 2011 Tohoku earthquake
Earth Planets Space
(2011) - et al.
Spatial distribution and focal mechanisms of aftershocks of the 2011 off the Pacific coast of Tohoku earthquake
Earth Planets Space
(2011)
Broad bandwidth study of the topography of natural rock surfaces
Journal of Geophysical Research
Three-dimensional simulation of spontaneous rupture: the effect of non-uniform prestress
Bulletin of the Seismological Society of America
The 2011 Tohoku-Oki earthquake: displacement reaching the trench axis
Science
Integral equation method for plane crack with arbitrary shape in 3D elastic medium
Bulletin of the Seismological Society of America
Interplate seismogenic zones along the Kuril–Japan trench inferred from GPS data inversion
Nature Geoscience
A new BIEM for rupture dynamics in half-space and its application to the 2008 Iwate–Miyagi Nairiku earthquake
Geophysical Journal International
Dynamic rupture propagation on a 2D fault with fractal frictional properties
Earth Planets Space
Striations, duration, migration, and tidal response in deep tremor
Nature
Earthquakes as multiscale dynamic rupture with heterogeneous fracture surface energy
Journal of Geophysical Research
Determination of constitutive relations of fault slip based on seismic wave analysis
Journal of Geophysical Research
Shallow dynamic overshoot and energetic deep rupture in the 2011 Mw 9.0 Tohoku-Oki earthquake
Science
Cited by (33)
A review of the rupture characteristics of the 2011 Tohoku-oki Mw 9.1 earthquake
2018, TectonophysicsCitation Excerpt :Multi-scale asperity models have been considered as representations of the Japan megathrust, accommodating the diverse range of large events and re-triggering of rupture areas of earlier events (Hori and Miyazaki, 2011). Aochi and Ide (2011, 2014) and Ide and Aochi (2013) compute spontaneous dynamic rupture in systems with variable circular patches with fracture energy proportional to patch size. Cascading ruptures occur in these models, with slow rupture in the high fracture energy large slip regions, similar to the 2011 event.
Study of Algorithms for Fast Computation of Crack Expansion Problem
2017, Procedia Computer ScienceRecurrence intervals for M > 7 Miyagi-ken-Oki earthquakes during an M ~ 9 earthquake cycle
2023, Progress in Earth and Planetary ScienceA review on slow earthquakes in the Japan Trench
2023, Progress in Earth and Planetary ScienceImpact of Hypocenter Location on Rupture Extent and Ground Motion: A Case Study of Southern Cascadia
2023, Journal of Geophysical Research: Solid EarthPaleotsunami history of Hachinohe, northern Japan: a multiproxy analysis and numerical modeling approach
2022, Progress in Earth and Planetary Science