IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.314315316.114.2009

Structural geology and geomechanics

Our interpretations of structure and in situ stress are derived from resistivity images in Hole C0002A (see the “Expedition 314 methods” chapter). We used a variety of images for the interpretation, including shallow, medium, and deep depths of investigation and several static and dynamically normalized images. Final interpretation was based on the shallow image. Data quality is good for the majority of the hole. See the “Expedition 314 methods” and “Expedition 314 Site C0001” chapters for comments on uncertainties and methods in picking features and in differentiating bedding and fracture planes. Where possible, the level of uncertainty of interpretation is provided in the text and figures of this chapter. Caliper data were not directly integrated with the resistivity images but were considered during interpretation.

Site C0002 penetrated both the forearc basin and the underlying older accretionary prism; therefore, the structural character of these two major units can be compared (Fig. F28). We also compare structural aspects of this site with Site C0001.

Bedding

Bedding dips contrast clearly between the forearc basin, with mostly shallow dips (≤15°), and the underlying prism, with markedly steeper dips (~30°–60°) (Figs. F28, F29). The shallow dips of the forearc basin make identifying and accurately determining dip and strike difficult, and this should be considered in the following interpretation. Our interpretation results, however, show that the poles to the bedding planes in the basin are highly concentrated and their trends generally agree with the regional bedding parameters defined by the three-dimensional (3-D) seismic data volume (Fig. F29), suggesting sufficient reliability. See “Discussion and synthesis” for comparisons of borehole image and seismic data.

The forearc basin is divided into three logging units (Figs. F28, F29). The majority of beds in logging Unit I trend northeast to east and dip to the south (<4°). Logging Unit II beds appear to trend approximately northeast–southwest but with some scatter in azimuth and dipping (<8°), both toward the north and south. Logging Unit III beds also trend northeast–southwest but are more variable in azimuth and dip magnitude, with dips to both north and south. Bedding dips increase in the lower half of logging Unit III (Fig. F30). For the structure of logging Unit III, see the next section.

The stratigraphy of the underlying prism (logging Unit IV) is more deformed with steeper dips (~18°–60°) and a dominant northeast–southwest trend (Fig. F29). The poles to the bedding of this unit have two clusters on the stereonet (Fig. F29B). One highly scattered cluster dipping to the south approximately corresponds to the shallow beds (~1120 m LSF), and the other less scattered one dipping to the north corresponds to the deeper beds.

Structure of logging Unit III (forearc basin–prism transition)

Logging Unit III is distinguished from other units by its lithologic homogeneity and distinct gamma ray and PEF signature and is interpreted as a homogeneous mud deposit (see “Log characterization and lithologic interpretation”). Structurally it also contrasts with logging Units II above and IV below (Fig. F30). Borehole breakouts, which occur throughout most of the hole, are very weakly developed within this unit, possibly as a result of sediment properties. The base of this unit (lower ~30 m) is characterized by a number of steeply dipping resistive fractures (Fig. F30).

Bedding is difficult to discern, but beds in the middle zone (870–910 m LSF) dip to the south at up to 25°, whereas beds in the lowermost 10 m of the unit dip to the north (<16°).

Natural fractures

Fractures were less common at this site compared with Site C0001 within the active accretionary prism. Fractures have been analyzed according to azimuth and conductivity and are classified into four types: (1) conductive, (2) resistive, (3) undefined conductivity, and (4) uncertain fractures. Most fractures only show a partial sinusoid in the borehole images, often preventing accurate determination of strike and dip. In some cases, these fractures have been classified as “uncertain” fractures (e.g., Fig. F28), as there is a small possibility that they are not real features. Uncertain fractures are not differentiated according to conductivity but are mostly conductive. In some cases, fracture conductivity cannot be defined because of small fracture aperture or little/variable resistivity contrast with surrounding sediments, hence the fracture class “undefined conductivity.” Fracture frequency (Fig. F31) indicates increased deformation in the prism (below 936 m LSF) relative to the forearc basin.

Fractures in the forearc basin are both resistive and conductive, but mostly conductive if all uncertain fractures are included (Figs. F28, F32C). In the fractured base of the forearc basin (logging Unit III), fractures are resistive (Fig. F30). The majority of better defined fractures within the prism (logging Unit IV) are also resistive but more mixed in conductivity if “uncertain” fractures are included.

Analyzing the entire hole, the fractures exhibit scatter in trend but with dominant trends of northeast–southwest and northwest–southeast (Fig. F32C). Fractures of the forearc basin (logging Units I–III) are highly scattered (Fig. F32A). However, trends of northeast–southwest to east–west may be present. Uncertain fractures show a clear northwest–southeast trend. A few of the forearc basin fractures, with orientations northeast–southwest and east-northeast–west-southwest, offset bedding planes with normal offset (Fig. F33). Higher fracture densities occur in the upper 200 m of the forearc basin and within logging Unit III (Fig. F31).

Prism fractures (logging Unit IV) are more difficult to differentiate from bedding but show much less scatter than the forearc basin and a dominant northeast–southwest trend with a minor northwest–southeast trend (Fig. F32B). Examples of conductive prism fractures are shown in Figure F34. Fracture dips are not significantly steeper than those in the basin (~30°–85°). Many of the fractures are bedding parallel within this unit. Several large-aperture (10–30 cm) highly resistive fractures were observed within the prism, and we interpret these as cemented or mineralized. These features were interpreted as fractures, not bedding, because of their extreme resistivity.

Borehole breakouts

Breakouts were common throughout Hole C0002A (e.g., Fig. F33) and present in both the forearc basin and underlying older accretionary prism. Interestingly, the breakout orientation is approximately the same in these two distinct units (Figs. F28, F35A). No tensile fractures were observed. The average azimuth of breakouts is ~135° (northwest–southeast) with a range of 080°–170°, indicating that SHmax is oriented northeast–southwest (045° or 225°). This orientation is at ~90° to that at Site C0001 in the megasplay fault zone or outer arc high. Breakout widths range from 10°–170° but with a modal range of ~20°–40° (i.e., at the lower end of the total range) and an average of ~60° (Fig. F35B).

In detail, breakout parameters (azimuth and width) vary slightly with depth (Fig. F36). Figure F36A illustrates azimuth varying nonlinearly downhole, generally from a mean of 120° in the upper forearc basin to a mean of 145° in the lower part of the prism penetrated by the borehole. The most dramatic change in orientation occurs at ~1200 m LSF from 125° to 140°. The general change in azimuth downhole is likely a result of changes in the stress field, but local variability on the 50–200 m depth scale may be related to sediment properties around the borehole. Breakout width (Fig. F36B) shows a distinct increase at ~1000 m LSF near the top of the old accretionary prism material, which might be related to physical properties of the borehole sediments and/or related drilling parameters (e.g., borehole pressure).

Stress magnitude analysis from breakout widths

A preliminary attempt to constrain stress magnitude based on breakout widths was carried out at two depths (900 and 1300 m LSF). These two depths were chosen because rock types were relatively well identified from multiple logs and the states of stress may be different at depths above and below the major forearc basin/​prism boundary. Average widths of borehole breakouts are 34° and 85° at the respective depths.

Formation pressure at each depth was assumed to be hydrostatic. Figure F37 shows stress polygons for the two depths, in which the drilling-induced tensile fracture criterion (indicated by a red line) and rock uniaxial compressive strength (UCS) lines (indicated by blue lines) are plotted. Each blue line represents a trace of SHmax and Shmin that is required to create the observed widths of borehole breakouts for a given UCS value. Because no drilling-induced tensile fractures were observed at Site C0002, SHmax and Shmin must lie within the stress polygon to the right of (below) the red line and along a blue line representing rock strength.

Reliable information on rock strength is essential for this analysis. Since no known rock strength data are available in this area, an indirect estimation of rock strength was attempted using a series of empirical relationships that relate velocity or porosity to UCS (Chang et al., 2006). These relations are calibrated for high-porosity shale and sandstones from the Gulf of Mexico and North Sea, which might have undergone a different compaction and diagenetic history from the rocks of the Nankai forearc. Thus, the estimated UCS values here are subject to uncertainty and await corrections using data from subsequent expeditions.

The values of UCS were estimated to be 5.1 ± 3.9 MPa (shale at 900 m LSF) and 11.9 ± 1.7 MPa (sandstone at 1300 m LSF). Based on these estimated UCS values, the state of stress at 900 m LSF is a normal fault stress regime (Shmin < SHmax < Sv) and that at 1300 m LSF is categorized to be either a strike-slip fault or a thrust fault stress regime.

Discussion and conclusions

Comparisons between borehole analysis of structure and seismic reflection data

Bedding in the 3-D seismic data set shows gentle dips in the Kumano forearc basin sediments, and this agrees with the structural style in the resistivity images (Fig. F29). Forearc basin bedding orientations from borehole data are generally compatible with those imaged by seismic reflection data (Table T6) with a northeast–southwest strike. Prominent seismic scale horizons in logging Unit I strike northeast–southwest and dip south, which agrees generally with small-scale bedding from the borehole. Logging Unit II beds strike northeast–southwest and dip mostly to the north in seismic data. The strike is consistent in borehole data, but dip directions are more scattered. This may be due to difficulties in resolving low-angle features or to some difference between bedding at two different scales.

Logging Unit III represents the lowermost forearc basin sediments and appears to be divided into two subunits in seismic reflection data (Fig. F14). The lower of these two subunits lacks continuous seismic reflectors and coincides with more intensive fracturing from resistivity images (Fig. F30). The two seismic units of this logging unit have contrasting dip directions: the upper dipping north and the lower dipping south. Borehole bedding interpretations (Fig. F30) show northeast–southwest trends and more scattered dip directions (north and south) but with a more consistent southeasterly dip direction and increased dip magnitude in the lower half of the unit (below ~870 m LSF), except for the basal 10 m where dips decrease again. This also generally agrees with the seismic data.

Within the prism, both seismic and borehole interpreted beds dip more steeply than the basin above and strike approximately northeast–southwest. The scattered azimuth and dip of the borehole beds suggest a strongly deformed structural style of the prism, which cannot be clearly imaged in the seismic profiles.

Seismic reflection data show clear normal faulting within the forearc basin (Figs. F14, F38) with trends from east–west to northeast–southwest and predominantly dipping to the north. There is another set of faults identified in time slices, with some displacing the seafloor, with a northwest–southeast trend. These seismic scale faults can be correlated with the forearc basin fractures identified in resistivity images, (i.e., those with northeast–southwest to east–west trends and northwest–southeast trends [uncertain fractures]) (Fig. F32A).

The orientation of possible faults within the upper prism has not been determined from seismic reflection data, although some of the northeast/east-trending normal faults in the basin may extend into and offset parts of the upper prism. If we presume that the tectonic setting has remained similar for the last few million years, then we might expect old compressional structures within the older prism unit with similar trends to the active prism (i.e., northeast–southwest), as observed in the borehole fracture data set at this site (Fig. F32B). It is possible that some of these structures may have been reactivated as other fault/fracture types in the present stress regime or, alternatively, these structures may be inactive prism fractures.

Stress magnitude implications for deformation style

Stress magnitude estimations from borehole breakout widths, compressive strengths, borehole pressures, and friction coefficients suggest different stress regimes within the forearc basin and prism (Fig. F37). At ~900 m LSF, the base of the forearc basin, normal faulting would be expected. In this case, the maximum principal stress would be vertical and the minimum principal stress would be to the northwest. This is generally consistent with the observed seismic scale faulting and with evidence of normal offset by borehole scale fractures trending ~northeast–southwest. At ~1300 m LSF, within the prism, the estimated state of stress conditions favor either strike-slip or thrust faulting. However, if the empirical equations have overestimated UCS, normal faulting is also possible. So the faulting stress regime is not well constrained in the older prism. Conceivably, the following deformation styles are geologically possible and plausible from the predicted state of stress in the prism:

  • Normal faulting and extension similar to the overlying forearc basin, with faults extending into the upper prism, possibly reactivating old thrust faults with favorable orientation;
  • Strike-slip faulting accommodating some component of margin-parallel deformation; and
  • Thrust faulting but with northwest–southeast fault trends (perpendicular to the margin) resulting from the orientation of SHmax.

The latter style of deformation, however, is not compatible with northeast–southwest borehole fracture orientations in the prism. If the strike-slip or thrust faulting stress regime dominates, then SHmax is also the maximum principal stress and trends parallel to the strike of the margin and prism. Ground-truthing by coring will reduce the uncertainty in the parameters for determining stress state and magnitude, which may allow these hypotheses to be tested.

Comparison of in situ stress and deformation between Sites C0001 and C0002

In situ stress indicators (borehole breakouts) suggest SHmax is rotated by ~90° at Site C0001 relative to Site C0002 located to the southeast, which is compatible with the northwest–southeast to north–south extension observed in the forearc basin. This abrupt change in stress regime over a short distance (~10 km) may represent the transition from the actively shortening prism to the relatively static landward part of the forearc. This corresponds well to the models of Fuller et al. (2006) and to some extent Wang and Hu (2006) that predict high strain rates within the active critical wedge (“outer wedge”) and relatively little deformation in the stable “inner wedge” (beneath and within a sediment-filled forearc basin). At the latter location, the convergence-related deformation is thought to be focused on the plate boundary at depth. The transition between the inner and outer wedges, which should be between Sites C0001 and C0002 based on deformation and morphology, is proposed to mark the approximate updip limit of the seismogenic zone (Wang and Hu, 2006). During the interseismic period, Wang and Hu (2006) predict compression in the inner wedge, which we do not observe. This may be due to the proximity of Site C0002 to the inner-outer wedge transition zone, shallowness of the borehole (so not sampling the true stress state of the inner wedge), variations in coefficient of friction along the plate boundary, or other problems with the model.

Extension within the forearc basin and potentially the underlying prism can be explained by gravitational collapse and downslope extension as movement on the outer arc megasplay fault zone causes uplift and backtilt of the seaward part of the basin (inner wedge). Another example of a forearc basin extending perpendicular to the margin is northern Cascadia, USA, where underlying overpressured prism melange drives downslope migration and extension of basin sediments (McNeill et al., 1997).

The gradual rotation of breakout orientations and therefore SHmax with depth at Site C0002 (Fig. F36) suggests that such a change could continue with depth to the megasplay fault or décollement where in situ stress might be expected to reflect plate convergence.