Stress State in the Largest Displacement Area of the 2011 Tohoku-Oki Earthquake

See allHide authors and affiliations

Science  08 Feb 2013:
Vol. 339, Issue 6120, pp. 687-690
DOI: 10.1126/science.1229379


The 2011 moment magnitude 9.0 Tohoku-Oki earthquake produced a maximum coseismic slip of more than 50 meters near the Japan trench, which could result in a completely reduced stress state in the region. We tested this hypothesis by determining the in situ stress state of the frontal prism from boreholes drilled by the Integrated Ocean Drilling Program approximately 1 year after the earthquake and by inferring the pre-earthquake stress state. On the basis of the horizontal stress orientations and magnitudes estimated from borehole breakouts and the increase in coseismic displacement during propagation of the rupture to the trench axis, in situ horizontal stress decreased during the earthquake. The stress change suggests an active slip of the frontal plate interface, which is consistent with coseismic fault weakening and a nearly total stress drop.

The huge tsunami associated with the 2011 Tohoku-Oki earthquake [moment magnitude (Mw) 9.0] was caused by the very large coseismic fault displacement of the shallow portion of the subduction zone near the Japan trench (15). Besides the unprecedented large coseismic slip of >50 m, the other surprising feature of the earthquake is that the large slip on the frontal plate interface reached the sea floor at the trench axis (6). The state of stress and frictional behavior of the frontal plate interface is important for controlling coseismic displacement. Indirect analyses on stress state change and/or stress drop associated with the 2011 Tohoku-Oki earthquake have been carried out from remotely sensed observations (712).

To investigate the stress change associated with the 2011 Tohoku-Oki earthquake, we analyzed geophysical logs collected by the Integrated Ocean Drilling Program (IODP) Expedition 343 (13, 14), Japan Trench Fast Drilling Project (JFAST), at site C0019 (Fig. 1, A and B), located ~93 km seaward from the epicenter of the mainshock and ~6 km landward of the trench axis. At this location, three boreholes successfully penetrated the interface between the subducting Pacific Plate and the overriding North American Plate. The boreholes enabled geophysical logging, core sampling, and long-term temperature monitoring (15).

Fig. 1

(A) Location of JFAST site C0019 and SHmax orientation in the deep part of the borehole. Red solid and dashed lines show the mean SHmax orientation and 1 SD, respectively. Green circles and lines show Ocean Drilling Program sites drilled in 1999 and their SHmax orientations before the 2011 earthquake (25). The red star denotes the epicenter of the 2011 earthquake mainshock. Yellow (31), black (32), and orange arrows (1) indicate coseismic horizontal displacements of sea floor, and the orange rectangle denotes the bathymetric survey area. The gray arrow shows relative plate motion (29). The white line around site C0019 shows the location of (B). (B) Interpreted seismic line crossing site C0019. The red vertical line shows the location and approximate drilled depth range. The subducting Pacific plate dips gently ~5°; the landward lower trench slope dips ~7°. CMP, common midpoint; mbsl, meters below sea level; V.E., vertical exaggeration. (C) Schematic of inferred coseismic three-dimensional stress state change in the lower portion of the frontal prism. σ1, σ2, σ3, and σv are the maximum, intermediate, minimum, and vertical stresses, respectively.

The JFAST borehole C0019B reached ~850 m below sea floor (mbsf) in a water depth of 6890 m and was used mainly to collect geophysical logging data. From borehole wall resistivity images, we observed clear borehole breakouts (16) in the depth range of ~44 to 813 mbsf (Fig. 2C). These drilling-induced compressive failures are reliable indicators of the orientations of current maximum and minimum horizontal stresses (SHmax and Shmin, respectively) and can be used to constrain stress magnitudes (17). C0019B penetrated through the plate-interface fault around 820 mbsf, the probable principal slip zone of the 2011 earthquake; no breakouts are observed below the fault (15). Above the plate interface, a Pleistocene accretionary prism consists of hemipelagic deposits lacking resolvable structural features on seismic reflection data (Fig. 1B) (1820). The frontal plate interface, beneath the accretionary prism, marks the transition to dominantly pelagic deposits.

Fig. 2

(A) Natural gamma-ray (GR) intensity (gAPI, a standard gamma-ray unit calibrated at the American Petroleum Institute) and (B) resistivity (Res). (C) SHmax azimuth determined from breakouts (red dots). Blue solid and dashed lines show the mean and SD, respectively, of SHmax azimuth in the deep part of the borehole (mean ± SD: 139 ± 23° or 319 ± 23°). (D) Widths of breakouts (WOB) (red circles) and their mean and SD in various units (blue solid circles and bars). Horizontal lines through (A) to (D) show the boundaries of the lithological units primarily defined by natural gamma-ray intensity (15). (E) Hydrostatic pressure and stress magnitudes [SV (black) was calculated from density; SHmax (red) and Shmin (blue) were constrained from breakout widths and rock strengths]. Thick black bars on the y axis show the depth range from where breakout data and UCSs constrained stress magnitude.

The SHmax azimuth and its variability differ markedly in the shallow and deep parts of the borehole (Fig. 2C). At shallow depths of ~44 to 197 mbsf within slope facies [unit I, derived mean porosity from logging-while-drilling resistivity and its standard deviation (SD) of ~62 ± 6% (15)], the SHmax varies from approximately parallel to the convergence direction to perpendicular (~100 mbsf) and back toward parallel again (~140 mbsf) (Fig. 2C). This suggests the presence of a discontinuity such as a fault, which is supported by changes in bedding dips and conductive peaks in the resistivity log (Fig. 2B). In the upper half of the wedge sediments from 197 to 537 mbsf [unit IIa, porosity of ~51 ± 3% (15)], the SHmax azimuth is highly variable. Such diverse breakout orientations have not previously been recognized at subduction and thrust-fault zones such as the Nankai Trough (2124), the Japan Trench margin before the 2011 earthquake (25), Costa Rica (26), and the Taiwan Chelungpu Fault (27, 28). Our observations suggest that SHmax and Shmin are close in magnitude, so that localized stress perturbations such as faults or topographic effects can be responsible for the scattered distribution. Correlation of the depth intervals, defined on the basis of SHmax azimuth distributions, and of the logging units, primarily defined by natural gamma-ray intensity, suggests some possible lithological influence on SHmax distributions (Fig. 2).

At the greater depths of ~537 to 813 mbsf [unit IIb, porosity of ~45 ± 3% (15)] within the accretionary prism and above the plate boundary, the SHmax azimuth has a clear preferred orientation in a northwest-to-southeast direction (319 ± 23°) (Fig. 2C). This stress orientation is consistent with the plate convergence direction of 292° (29) and also roughly consistent with stress orientations determined at sites 1150 and 1151 (drilled in 1999 before the 2011 earthquake) (Fig. 1A) (25). In a similar study of the Chelungpu fault that ruptured during the 1999 Chichi, Taiwan earthquake (Mw 7.6, thrust focal mechanism), two boreholes that penetrated the fault were drilled about 5 years after the earthquake. Observations of borehole breakouts after this incident indicate that the SHmax azimuth orientation changed by 90° in the vicinity of the fault (27, 28). Although a similar change in stress directions was not observed across the plate interface at ~820 mbsf in the JFAST hole, the analysis described below shows a large change in the stress state before and after the earthquake. The absence of breakouts below ~820 mbsf may be indicative of a change in stress state across the plate interface (Fig. 2).

The magnitudes of SHmax and Shmin are constrained from observed widths of breakouts (WOBs) and measured unconfined compressive strengths (UCSs) for two depth intervals around 720 and 812 mbsf, where both the WOBs and UCSs could be measured (16). We assume Andersonian stress states and a vertical stress SV calculated from a sediment density profile derived from the logging data (15): The values of SV, SHmax, and Shmin at 720 mbsf are approximately the maximum, intermediate, and minimum principal stresses (σ1, σ2, and σ3), respectively. At 812 mbsf, however, there is some uncertainty if SHmax is necessarily less than SV; the constrained stress state is close to the boundary of the normal faulting and strike-slip faulting regimes. Overall, these results indicate that the postearthquake stress states in the frontal prism are either in or close to the normal faulting stress regime (Fig. 2E).

Assuming that SV, SHmax, and Shmin are the three principal stresses, the current shear stress on a plane at 812 mbsf parallel with the plate interface at ~820 mbsf (for thrust motion in the convergence direction) can be calculated from SV, SHmax, and the dip of the interface (~5°). Using the greatest estimated value of SHmax (87 MPa), the shear stress on the plane at 812 mbsf after the earthquake is very small, less than ~0.3 MPa; the shear stress on the nearby plate interface can be inferred to be similar. In addition, we infer that decreases of SHmax, Shmin, and the mean stress of three principal stresses above the ~820-mbsf plate interface are around 2, 1, and 1 MPa, respectively, on the basis of Andersonian stress states before and after the earthquake, coseismic slip distribution near the trench, and elastic elongation of the frontal prism during the earthquake (16).

Unconfined compressive strengths of rocks in the range of ~800 to 820 mbsf (the bottom of unit IIb) are higher than those in the 821- to 836-mbsf range (unit III, sheared brown claystone) (16). Thus, the presence of breakouts in the stronger interval above the interface (and not in the weaker interval below the interface) is noteworthy. In general, for horizontal stresses of similar magnitude, breakouts occur more easily or are wider in weaker rock. Therefore, we believe that the horizontal stress (especially SHmax) just below the interface (unit III) is lower than that above the interface (the bottom of unit IIb). Namely, the magnitude of horizontal stress appears to abruptly change across the 820-mbsf fault.

The frontal portion of the accretionary wedge is considered to have been under trench-normal compression before the earthquake, on the basis of the orientations of minor faults and bedding observed in the logging data and in core samples. The faults and bedding are variable in dip magnitude, but faults at depths greater than ~690 mbsf display predominantly reverse shear sense. Bedding at all depths in the prism shows a preferred northeast strike direction reflecting horizontal contraction and local extension at shallower depths, approximately parallel to the plate convergence direction (15). These observations show that the overall structure in the hanging wall of the frontal interface is characterized as folding and thrusting. Also, a comparison of seismic reflection data before and after the Tohoku-Oki earthquake, from just 15 km north of the JFAST drill site (the orange rectangle in Fig. 1A), reveals that deformation at the trench axis formed as a result of compression during coseismic slip on the shallow plate interface (6). Therefore, we conclude that the stress state in the frontal prism has changed from a thrust faulting regime before the earthquake to the present normal faulting, or near-normal faulting regime. The present SHmax is approximately parallel to the plate convergence direction and is inferred to have changed from the maximum to intermediate principal stress (Fig. 1C) due to the nearly complete shear-stress drop. Most earthquakes are thought to have partial stress drops, but the results of this study and other indirect measurements or modeling (2, 7, 8, 30) suggest that the change in the stress state associated with the very large slip in this region is an indication of coseismic fault weakening and a nearly total stress drop for the Tohoku-Oki earthquake. This finding is consistent with increased coseismic displacement of the sea floor toward the trench axis, as observed in changes of bathymetric data and rupture models obtained from the inversion of seismic, geodetic, and tsunami data (Fig. 1A) (1, 3, 5, 3133).

Supplementary Materials

Materials and Methods

Supplementary Text

Figs. S1 to S5

Table S1

Reference (34)

Expedition 343 Scientists

Becky Cook,1 Tamara Jeppson,2 Monica Wolfson-Schwehr,3 Yoshinori Sanada,4 Saneatsu Saito,4 Yukari Kido,4 Takehiro Hirose,4 Jan H. Behrmann,5 Matt Ikari,6 Kohtaro Ujiie,7 Christie Rowe,8 James Kirkpatrick,9* Santanu Bose,10 Christine Regalla,11 Francesca Remitti,12 Virginia Toy,13 Patrick Fulton,14† Toshiaki Mishima,15 Tao Yang,16 Tianhaozhe Sun,17 Tsuyoshi Ishikawa,4 James Sample,18 Ken Takai,4 Jun Kameda,19 Sean Toczko,4 Lena Maeda,4 Shuichi Kodaira,4 Ryota Hino,20 Demian Saffer11

1University of Southampton, Southampton, UK. 2University of Wisconsin-Madison, Madison, WI, USA. 3University of New Hampshire, Durham, NH, USA. 4JAMSTEC, Yokosuka, Japan. 5GEOMAR, Kiel, Germany. 6University of Bremen, Bremen, Germany. 7University of Tsukuba, Tsukuba, Japan. 8McGill University, Montreal, Quebec, Canada. 9University of California, Santa Cruz, CA, USA. 10University of Calcutta, Calcutta, India. 11The Pennsylvania State University, University Park, PA, USA. 12Universitá di Modena e Reggio Emilia, Modena, Italy. 13University of Otago, Dunedin, New Zealand. 14University of Texas, Austin, TX, USA. 15Osaka City University, Osaka, Japan. 16Institute of Geophysics, China Earthquake Administration, Beijing, China. 17University of Victoria, Victoria, British Columbia, Canada. 18Northern Arizona University, Flagstaff, AZ, USA. 19University of Tokyo, Tokyo, Japan. 20Tohoku University, Sendai, Japan.

*Present address: Colorado State University, Fort Collins, CO, USA.

†Present address: University of California, Santa Cruz, CA, USA.

References and Notes

  1. Material and methods are available as supplementary materials on Science Online.
  2. Acknowledgments: For our research, we used data provided by IODP. The data will be distributed by IODP ( We thank all drilling and logging operation staff on board the D/V Chikyu during expedition 343. We gratefully acknowledge two anonymous reviewers for their constructive comments, which helped us to greatly improve this manuscript. Part of this work was supported by grant KAKENHI 22403008 (Japan Society for the Promotion of Science), grant 21107006 (Ministry of Education, Culture, Sports, Science and Technology of Japan), and the U.S. Science Support Program of IODP.
View Abstract

Navigate This Article