To understand the stress regime in the shallow portion of the Nankai Trough seismogenic zone, we carefully investigated the resistivity image logs at IODP Site C0009 that penetrated to 1600 meters below the sea floor (mbsf) in the central Kumano forearc basin. During the riser drilling, several downhole measurements were run including image logs, caliper and comprehensive geophysical logs sets. Borehole resistivity images obtained by a wireline-logging tool were reprocessed in order to eliminate image artifacts generated by sticky tool movements, etc. The constraints on the possible magnitude and orientation of horizontal principal stresses are provided from the borehole images, rock strength, and logging data. The absence of borehole breakouts above 1285 mbsf represents a small difference (<10 MPa) between horizontal principal stresses and reflects the normal faulting and strike–slip faulting stress regimes. A stable, continuous breakout was observed in the resistivity image logs suggest that the stress state is in a strike–slip faulting and tending to reverse faulting regime below 1285 mbsf. The differential horizontal stress (up to 15 MPa) is significantly larger than in the shallow portion of the borehole (SHAMX ≫ Sv > Shmin).
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
 Nankai Trough Seismogenic Zone Experiments (NanTroSEIZE) comprise one of the projects of the Integrated Ocean Drilling Program (IODP). It aims to sample and monitor the seismogenic zone and the shallow portion of the subducted plate boundary, offshore southeastern Japan (Figure 1a). From 2007 to 2009, on-site stress orientations were determined from borehole breakouts and drilling-induced tensile fractures identified from logging while drilling (LWD) image logs in six holes [Tobin et al., 2009; Saffer et al., 2010; Chang et al., 2010]. Site C0006 was drilled near the toe across the frontal thrust; Sites C0001, C0010, and C0004 were drilled near the megasplay fault (Figure 1b). The horizontal maximum principal stress (SHMAX) in these four boreholes is oriented to the sub parallel direction (N36°W) with the plate movement vector of Philippine Sea plate (N55°W [Miyazaki and Heki, 2001]). Site C0002 is located near the shelf break, with the SHMAX rotated approximately 90° compared to other drill sites (Figures 1a and 1b) [Yamada et al., 2011].
 In 2009, IODP Expedition 319 was carried out to drill in the central Kumano basin through the forearc basin and probably into the early accretionary prism from 700 to 1600 mbsf [Saffer et al., 2010]. The scientific objectives included characterizing the lithology, physical and chemical properties of the faults, and the strain accumulation in the active plate boundary. Comprehensive wireline logging operations were carried out in this borehole, such as P-wave and S-wave velocities and borehole resistivity images, as well as other basic information (e.g., density, gamma ray, resistivity, calipers, temperature, mud weight, and mud gas).
2. Borehole Condition and Stress Orientation at Site C0009
 Many drilling projects and expeditions such like Cajon Pass well; TCDP and NanTroSEIZE Expedition 314 have indicated that the borehole breakout and tensile fractures are formed by the acting principal stresses [Moos and Zoback, 1990; Wu et al., 2007; Lin et al., 2010]. High-resolution resistivity images measured from Formation MicroImager (FMI) are widely used for detection of such stress estimation. The FMI tool consists of two pairs of orthogonal pads (C1 and C2), which measure the borehole diameter for the whole surveyed depth range. As shown inFigure 2a, the caliper log (C1 and C2) does not provide evidence for significant breakouts between 700 and 1285 mbsf, which indicated that the shape of the wellbore is in gauge hole (C1 = C2 = BZ) during this section. Below the 1285 mbsf, two borehole breakout zones were identified, caliper one (C1) locks into breakout zone from 1285–1460 mbsf, the tool then rotated 90° and caliper two (C2) locks into another breakout zones from 1460–1560 mbsf. Both zones indicate the azimuth of breakout in the borehole that deviated angle is less than 0.25° (Figures 2a and 2b). These elongations correspond to cease the tool rotation in the zone of borehole enlargement [see, e.g., Plumb and Hickman, 1985]. The washout zone was reported in the depth 1560–1580 mbsf which implied the borehole enlargement at the all azimuth. Except the washout zone, the well-round borehole shows the stress concentration on the borehole wall is under the strength of formation. The one direction enlargement of breakout acquired from the Pad azimuth indicating the orientation of Shmin. Since the borehole deviation of C0009 is kept at less than 0.25 degrees for most of its interval (Figure 2c), the hole deviation effect on the azimuth of breakouts and tensile cracks can be neglected in this analysis.
 Four lithology units were defined based on cutting and wireline logging (Figure 2d). Silty mud with sand-rich layers of Pleistocene-Holocene age is identified at 0–791 mbsf (Unit I and II) while Unit III (791–1285 mbsf), dated from the Pleistocene to the late Pliocene, consists of silty clay with higher wood content. Below 1285 msbf, the silty mudstone from the late Miocene was recognized as Unit IV [Saffer et al., 2010].
Saffer et al.  and Lin et al. reported there is no breakout identified shallower than 1285 mbsf. However, the stick and slip motion of the tool due to the rugose or sticky borehole condition distorted and blurred the initial image which recorded incorrect depth and distorting images. We applied the tool motion correction (tool velocity, cable speed, acceleration on Z axis and tool head tension) to improve FMI image qualities. The reprocessed data indicated no drilling-induced tensile fractures observed in the entire borehole, in contrast to the reports ofLin et al. . Single vertical fractures (SVF) possibly scratched by drilling bit were discovered in the shallow depths of borehole, which recognized as tensile fractures. However, SVF probably did not reflect the stress direction around the borehole wall (Figure 2e).
 In general, the absence both of breakout failure and tensile fractures in this section provides no information on the stress orientation above 1285 mbsf. Below 1285 mbsf (Unit IV), the breakouts longer than one meter were examined. The breakout azimuth is basically consistent throughout the interval (Figure 2d), and the direction of Shmin is N54 ± 9°E, which is highly consistent with estimates by Saffer et al.  and Lin et al. . The total length of 181.3 meters in over 129 breakouts were observed with the low standard deviation (s.d. = 9°) and measured in this wellbore assigning an A-quality factor according to World Stress Map ranking criteria (O. Heidbach et al., The World Stress Map database release, 2008, doi:10.1594/GFZ.WSM.Rel2008). The azimuth of SHMAX converted from borehole breakout has the direction of N36 ± 9°W, which is rotated by 19 degrees from the orthogonal direction of tectonic plate motion by oceanic continental plate convergence vector [Heki, 2007].
3. Estimation of the Magnitude of the Horizontal Principal Stress
 To estimate the magnitude of horizontal principal stresses (Shmin and SHMAX), we followed the methodology that requires rock strength (CO) and logging data to constrain the borehole stress state [Moos and Zoback, 1990; Hickman and Zoback, 2004]. The survey from cutting, cores and wireline logging by Saffer et al. reported that the lithology units in this borehole are mudstone with sand-rich layers. However, the rock-mechanical properties measured for the core samples from Site C0009 is not available due to less rock experiments in this borehole. Thus we adopted the empirical relationships between Vp and CO for sandstones (Figure 3a). The upper bound of the rock strength can be considered by Freyburg :
where CO is in MPa and Vp is in m/s. However, the wide ranges of rock strength estimation are difficult to constrain the minimum required rock strength. The effective stress acting on the borehole wall needs the further discussion in following.
 The stress concentration around the borehole wall can be described in terms of the principal stress components. This concentration varies strongly as a function of the azimuth around the wellbore [Zoback, 2007] as showing in Figure 3b. The effective stress acting on the borehole wall varies with the azimuth in a sinuous curve, so called the hoop stress (σθθ). The maximum and minimum hoop stress are described as following by Barton et al. , Jaeger et al. , and Haimson et al. :
where θ is the angle measured from the azimuth of SHMAX, P0 is the pore pressure, ΔP is the difference between the borehole pressure (mud weight, Pm) and P0. σΔT represents thermal stresses arising from the difference between the mud temperature and formation temperature, and 2θb is the angle related to the width of breakout. The compression is taken as positive for all stress and pressure in these equations. We neglect the thermal stress term in this study, for no significant temperature anomaly was observed in this drilling (Figure 2c).
 We are able to constrain the stress parameters by the direct observation of geophysical logs. The vertical stress (Sv) and hydrostatic pore pressure (P0) is calculated from the averaged density log (2039 kg/m3) and the density of seawater (1030 kg/m3), respectively. In order to stabilize the borehole condition, weight-adjusted drilling mud was used, and the mud pressure (Pm) was shown to be 3–4 MPa above the pore pressure by the static mud pressure measurements [Saffer et al., 2010]. From the hydraulic fracturing (HF) tests at Site C0009, there are two magnitudes of Shmin reported at depths of 873.7 and 1532.7 mbsf and more confidence in the experiment at the shallower depth [Saffer et al., 2010], the magnitude of Shmin measured from HF test is 35 MPa reported by Ito et al.  at 873.7 mbsf.
 Since no tensile fractures were observed in the shallow part (<1285 mbsf), the minimum hoop stress σθθmin is larger than zero (σθθmin ≥ = −T0 ∼ 0). The maximum magnitude of SHMAX is 41 MPa calculated by equation (2) in known Shmin. The stress state in the borehole must lie within the bounds imposed by the definition that SHMAX ≥ Shmin [Moos and Zoback, 1990]. The maximum value of σθθmin can be constrained by setting the stress state in the homogeneous stress case (SHMAX ≈ Shmin). In this case, equation (2) derived to
The magnitude of Shmin was constrained by HF test; the upper bound of σθθmin is around 8 MPa and constrained the range of σθθmin between 0−8 MPa. Based on these calculated SHMAX and Shmin (41 MPa and 35 MPa) above, the magnitude of σθθMAX at this depth is verified by equation (3) equal to 24 MPa. This value was marked as bearing the lower bound of the rock strength at 873.7 mbsf where we regressed from the by Lama and Vutukuri  and Carmichael  data (Figure 3a; blue triangle at 873.7 mbsf). This information revealed that the possible range of σθθMAX can be set to the corresponding Co from 24 to 45 MPa estimated from the equation (1), which showed 21 MPa difference. Consider the both limitations of σθθmin (0−8 MPa) and σθθMAX (UCS − 21 ≤ σθθMAX ≤ UCS), we applied these boundary conditions of hoop stresses to calculate the each SHMAX and Shmin in every depth of shallow part by equations (2) and (3).
 In the deeper part (below 1285 mbsf), if we use the value of Shmin in 1532.7 mbsf measured by HF tests for hoop stress estimation (SHMAX ≈ Shmin), σθθmin was negative and the drilling induced tensile fractures (DITF) took place below 1285 mbsf. However, we did not recognize any DITF from the image logs. This result implied that the magnitude of Shmin, 41 MPa, measured from HF tests is under-estimated the magnitude of Shmin. For satisfying the observation of no tensile fracture in FMI logging, we set the σθθmin level to 0 (0 ≈ σθθmin) which means the larger value of Shmin we expected at this depth (Shmin = 45 MPa) by equation (5). The magnitude of SHMAX can be calculated with related Shmin in equation (2) between 54–59 MPa showing in Figure 4 at depth 1532.7 mbsf. In this scenario, the generated maximum hoop stress (σθθMAX) showing in Figure 3a at 1532.7 mbsf (blue triangle) using equation (4), with the maximum breakout width (52°) we can confirm from FMI. This magnitude indicated the lower bound of σθθMAX and its possible range comparing to the rock strength estimation is 42 MPa, which is 13 MPa apart from the upper bond of rock strength (55 MPa). Rely on the knowing range of σθθmin (close to zero) and σθθMAX (UCS − 13 ≤ σθθMAX ≤ UCS) in the deeper part, the pairs of principal stresses (SHMAX and Shmin) can be obtained with every depth calculating by equations (2) and (4) below 1285 mbsf. The upper and lower boundaries of rock strength can be seen in Figure 3c.
 The principal stress profiles are plotted in Figure 4 with depth, which calculated by constrained hoop stress and rock strength mentioned in previous paragraphs. These curves suggest that the stress regime can be categorized into three segments: In the shallow part, Shmin is close to but lower than Sv. For example, at the depth 873.7 mbsf with Sv = 39 MPa, the magnitude of SHMAX is 42 to 51 MPa and the Shmin is around 38 ± 3 MPa. The horizontal differential stress is generally small (<10 MPa). In the accretionary prism in Unit IV (deeper than 1285 mbsf), Shmin remains lower than Sv, and SHMAX stays higher than Sv. The stress state at depth 1532.7 mbsf shows that Sv = 52 MPa, Shmin = 47 ± 2 MPa, and SHMAX is greater than 59 to 64 MPa (estimated from the breakout width 52°). We suggest that the horizontal differential stress is around 15 MPa.
Saffer et al.  and Doan et al. reported the segment of gas zones above 1285 mbsf. Low P-wave velocity is characterized in these sections. The magnitude of stresses in the gas zones showing inFigure 4 display the lower values of horizontal principal stresses, in which Shmin is 39 ± 2 MPa < Sv (45 MPa) ≅ SHMAX (41–47 MPa) at the depth 1150 mbsf. The lowest horizontal differential stress (SHMAX − Shmin) were pointed out in gas zones. The low differential stress suggests that the gas zones have less anisotropy, which indicates the stress state in the normal fault stress regime and is consistent with the gas zones at a low Vp/Vs ratio.
4. Discussion and Conclusions
 Due to the lack of the on-site rock experiments, there is the uncertainty for the rock strength estimation. The empirical functions from UCS may not reflect the real type of the rock or on-site rock strength. However, the calculation of hoop stresses in this study indicated that the maximum hoop stresses are significantly consistent with the range of rock strength we assumed. It is reasonable to use these empirical functions to constrain Co under the hydraulic fracturing test results, conditions of tensile fracture and breakout width.
 In the forearc basin (700–1285 mbsf), the estimated range of stress magnitude suggests the normal faulting or transition to the strike–slip regime, with a less anisotropic stress situation (SHMAX >≈ Shmin ≈ Sv). The stress state in the accretionary prism (1285–1600 mbsf) transfers to the strike–slip regime and suggests higher amount of anisotropy due to the larger differential horizontal principal stress about 15 MPa. This unit displays a uniform borehole breakout azimuth at N54 ± 9°E (SHMAX is oriented at N36 ± 9°W). From the core observation and 3D seismic data, there is no significant stress perturbation at site C0009 [Saffer et al., 2010]. This suggests that the orientation of SHMAX in this borehole would follow the regional tectonic stress direction. The possible range of stress magnitude indicates the strike–slip faulting regime or a transition to the normal fault regime in the forearc basin.
 At Site C0002, the SHMAX direction is oriented perpendicular to the plate convergence [Tobin et al., 2009]. Such differences stress orientation between Sites C0009 and C0002 would be attributed to the local, margin-perpendicular gravitation-driven extension overrides the regional tectonic compression around Site C0002. The single stress orientation varies such like the 90° rotation of breakout azimuth in the different geologic domains (near fault zones or Trough) from nearby sites (Figure 1a) also consistent with the report from Liu et al. .
 We grateful thank JAMSTEC, IODP, and Expedition 319 onboard scientists. The quality of this paper was greatly enhanced by six anonymous reviewers. This research was funded by Grant-in-Aid for Scientific Research on Innovative Areas (21107006) MEXT, Japan.
 The Editor thanks two anonymous reviewers for assisting in the evaluation of this paper.