Imaging of the borehole by RAB provided one of the immediate products of Leg 196 (Figs. F7, F8). The RAB images provide an in situ image of borehole structure unparalleled in completeness, although at a lower resolution than the cores. Nevertheless, combining these images with information on core structures provides probably the best view so far of the initial deformation in a subduction zone. At Site 1173 we imaged steeply dipping normal faults and fractures of variable strikes, which would be expected from the state of stress in a consolidating sedimentary basin. Conversely, the fractures are more consistent in dip azimuth in Hole 808I (Fig. F7), suggesting an influence of the compressional deformation at the frontal thrust. One of the most spectacular results of Leg 196 is imaging the breakouts in Hole 808I and their definition of a principal stress orientation parallel to that determined by the inversion of small faults (Lallemant et al., 1993) and parallel to that expected from the convergence vector (Seno et al., 1993). These breakouts are concentrated in lithologic Subunit IIC at Site 808. Additional analysis of breakout orientation, width, and depth along with determination of the cohesion and coefficient of friction of the sedimentary rocks should provide a better resolution of stress orientation as well as stress magnitude.
Tectonic consolidation of sediment in fault zones is influenced by fluid pressure. Thus, depth profiles of density or porosity variation through fault zones provides an overall view of fault zone behavior and a qualitative predictor of fluid pressure. From Hole 808I, we collected log data through the frontal thrust and décollement zones. This information, combined with existing core-scale density and porosity information, provides a unique view of major fault dynamics.
The density and density-derived porosity logs are suspect in the frontal thrust and décollement zones because of enlarged hole diameter. However, resistivity data provide insights on porosity variation. For example, the frontal thrust zone shows a sharp and sustained increase in resistivity. Because the pore water shows no significant variations in composition through this zone, it is unlikely that the resistivity increase is caused by differing fluid composition. Rather, the increase in resistivity may indicate a densification of the rock unit due to compaction accompanying shear.
At the top of the décollement zone, the resistivity trend changes from gradually increasing to gradually decreasing. The pore water chemistry around the décollement zone shows no anomalies that would explain the decreasing resistivity. The decreasing resistivity trend is presumably a response to the increased amount of fluid-filled unhealed fractures and a bulk porosity increase. In contrast, core porosity decreases in the décollement zone. Therefore, a combination of log and core measurements suggests that the décollement is an interval of enhanced porosity (probably fracture porosity) that encompasses blocks of sediment of relatively lower porosity and high density. Thus, the log data suggest that the fracture porosity of the décollement zone is dilated, probably held open by high fluid pressure, in contrast to the frontal thrust zone that is densified and may not be currently as highly overpressured.
The interpretations of fault zone porosity based on qualitative interpretation of resistivity must be verified by calculation of resistivity-derived porosity and appropriate corrections to the density data so that they can be utilized to calculate porosity. Additionally, direct fluid pressure measurements from the ACORK installation may provide information on fault zone pore pressures.
At Site 1173 density and porosity change downsection atypically for a normally consolidating sedimentary basin (Fig. F5). After remaining constant for ~200 m above, the porosity decreases sharply at ~340 mbsf and follows a normal consolidation curve below (Shipboard Scientific Party, 2001a). The porosity decrease at ~340 mbsf is associated with a diagenetic shift from cristobalite to quartz (Shipboard Scientific Party, 2001a), probably due to the alteration of vitric volcanic ash (Tada and Iijima, 1983) and siliceous microfossils. The stepwise porosity reduction at the cristobalite–quartz transition at Site 1173 mimics similar sharp porosity reductions across a similar phase transition (opal-CT to quartz) in siliceous rocks (Isaacs et al., 1983). Isaacs et al. (1983) believe that the opal-CT to quartz transition is associated with a reorganization and partial collapse of the sediment fabric that had previously been held open by an opal-CT cementation effect. A similar process could explain the sharp shift in porosity at Site 1173. At Site 1173 the abrupt shift in porosity and density at 320 to 360 mbsf generates a strong seismic reflection. This reflection cuts upsection across stratigraphy from southeast of Site 1173 northwesterly toward the margin (Fig. F4). An important question is, "What is the three-dimensional geometry of this reflector?"
The cristobalite to quartz phase transition occurs above the stratigraphic level to which the déollement would project at Site 1173, also at Site 1174, and, arguably, at Site 1177. Thus, the décollement seems to propagate seaward entirely in the apparently stronger, more dewatered portion of the Shikoku Basin formation and does not jump upsection to the less-dewatered section above 320 mbsf. Key questions raised by this observation are, "What is the real strength difference between the sediments above and below the cristobalite to quartz phase transition?" and "What is its effect on décollement development?"
For the first time in ODP history, LWD sonic velocity data were recorded during Leg 196 using the Schlumberger ISONIC tool. In situ velocity information is a key measurement because velocity is the fundamental link between seismic imaging and geology and because velocity is highly sensitive to stress, limiting the utility of shipboard core sample measurements. ISONIC data were recorded at both Sites 1173 and 808. Owing to the complex signal arrivals and the variety of phases propagating along the tool, picking compressional wave traveltimes (and, hence, velocity) is not straightforward. Initial traveltime analysis of the waveforms recorded during the leg did not produce reliable velocity values, although features in the velocity logs broadly correlate with those in the resistivity and density logs. Usable velocity and acoustic impedance results will require detailed waveform analysis to be carried out postcruise.