IODP

doi:10.2204/iodp.pr.316.2008

Discussion

As a result of the expedition, several unsolved questions were clarified although results and synthesis are still in a preliminary stage. We briefly discuss these preliminary results in this section.

Activity of the megasplay fault

Seismic and tsunami inversion studies combined with seismic reflection research strongly suggested that the splay fault drilled at Site C0004 slipped during the 1944 Tonankai M 8.2 earthquake (Park et al., 2002) and generated a tsunami. Moore et al. (2007) also suggested recent historical and geological accumulation of displacement along the fault on the basis of 3-D seismic reflection investigation. Although various lines of evidence suggest relatively recent activity along the splay fault, there is one conflicting feature on the seismic profile. The splay fault clearly thrusts hanging wall prism over younger slope sediments in the footwall; however, the youngest slope sediments that cover the fault appear not to be cut by the fault. In addition, the lack of a slope break on the seafloor above the fault could also suggest that this splay fault is presently not active but ceased activity in the recent past. However, shipboard results provide support for an alternate interpretation: the splay fault is active as a kind of blind thrust in which the tip of the fault has not propagated to the surface but remains buried. Slip on deeper levels of the fault zone are expressed by some combination of folding and layer-parallel slip in the shallow slope sediments draping the uppermost portion of the megasplay, as well as the possible triggering of marine slump or slide deposits.

Drilling results at Site C0004 indicate that the shallowest cover sediments above the hanging wall wedge are composed of repeated mass flow deposits associated with repeated slope collapses and rip-up debris generation. Pleistocene cover sediments dip steeply approximately parallel to the slope and are cut by numerous normal faults. Normal faults are also developed within the slope basin sediments in Holes C0008A and C0008C. These lines of evidence document that the slope is quite unstable and easily collapses when triggered, possibly by earthquakes. The age reversal from Pliocene to Pleistocene documented beneath the splay fault is also consistent with geologically recent activity.

Despite inferences of recent activity on the splay fault system, no porosity inversion is observed beneath the splay fault; this contrasts with previous results from the décollement of the Muroto transect (Screaton et al., 2002), in which a clear porosity inversion across the fault likely reflects fluid overpressure. Unlike the Muroto transect décollement, the splay fault system observed at Sites C0004 and C0008 has permeable pathways for dewatering provided by the observed sand and coarse ash layers.

Sediments of the slope basin at Site C0008 provide a "reference site" for the sediments underthrust beneath the megasplay fault. Comparison of the interval 190–200 m CSF in Hole C0008A with an average porosity of 50% and the correlated interval 320–330 m CSF in Hole C0004D with an average porosity of 43% suggests the sediments are dewatering during underthrusting. Evidence for lateral flow is provided by C1/C2 ratios at Site C0008 that are slightly lower than expected for biogenic production at the estimated in situ temperature. Lateral flow along sand layers could transmit fluids from where the temperature is higher because of greater burial underneath the splay fault.

These sand layers are truncated at a normal fault drilled in Hole C0008C, where surficial material has slid downward. Hole C0008C structural descriptions document normal faults within the sediments at ~40 m CSF. Lateral transmission of fluids from areas with thicker to thinner overburden has previously been suggested as a mechanism for enhancing slope failure (Dugan and Flemings, 2000). As a result, splay fault movement could produce slope failures through seaward propagation of pore pressures from the footwall in addition to oversteepening of the hanging wall. Postcruise examination of lithologic evidence, hydrogeologic properties, and geotechnical properties will help assess the interaction between splay fault movement and slope failures.

Permeable pathways for consolidation and dewatering of the underthrust sediments could have implications for splay fault evolution. As these sediments dewater, drainage will affect the profile of effective stress in the section, resulting in a migrating zone of minimum mechanical strength. As this progresses through time, it is possible that the splay fault will migrate to less consolidated sediments (Moore and Byrne, 1987; Saffer, 2003).

Two steps of age reversal are tentatively recognized across the splay fault zone; this evidence suggests that fault-bounded Unit III at Site C0004 is a sliverlike unit coming up from a much deeper setting. The lithology of the slope sediments and the old accretionary prism in Hole C0008A suggests that one of the possible sources for Unit III "sliver" at Site C0004 is the lowermost slope sediments beneath the Pleistocene and late Pliocene slope sediments. In this case, displacement along the splay fault, especially the lower boundary fault beneath Unit III, might be more than a couple of kilometers, which would require hundreds of large earthquakes, assuming meter-scale coseismic slip in this location (Tanioka and Satake, 2001; Kikuchi et al., 2003). Interestingly, porosities within this lithologic unit are slightly higher than expected relative to trends observed in overlying material. If these materials have been brought up from depth, they either never had an opportunity to consolidate or have subsequently had considerable opening of porosity by microfractures in this fault sliver.

Detailed analysis of the fault zone, including properties of the hanging wall, fault sliver, and footwall, is a main target of postcruise science to clarify the fault mechanism in the shallow splay fault.

Accretionary prism and plate boundary frontal thrust

The accretionary prism and frontal thrust cored during Expedition 316 (Sites C0006 and C0007) present unique features that are not observed in the Muroto and Ashizuri transects in the Nankai Trough or at other accretive margins such as Barbados, Aleutian, or Cascadia margins.

The features are

  1. Systematic in-sequence thrust propagating mode within the accretionary prism, which is well observed in the Muroto transect and the Barbados Ridge, is not observed; rather, the frontal thrust appears to have functioned as a plate boundary thrust for a long period, similar to the architecture observed along erosive margins.

  2. Taper angle at the toe is the largest in the Nankai Trough (Kimura et al., 2007a) and similar to those in erosive margins (Clift and Vannucchi, 2004).

  3. Seismic reflection profiles suggest that a new incipient décollement appears to be formed within a layer beneath the trench wedge sediments (G. Moore, pers. comm., 2007)

  4. Sandy and channel-filling coarse sediments are interpreted based on geophysical evidence to constitute most of the accretionary prism, and the same sediments fill the present trench wedge in the Nankai Trough.

How these characteristic features of accretionary prism in this region are intimately related to each other is unclear.

As a result of drilling at two sites (Holes C0006E, C0006F, C0007A–C0007C, and C0007D), several new aspects of the accretion/subduction system have been revealed. The upper accretionary prism is composed of coarse terrigenous sediments, including gravel-dominated and lower mud-dominated lithologies in the upper and lower parts of the accretionary wedge, respectively. They may be trench wedge to slope sediments and hemipelagic Shikoku Basin sediments, respectively; however, that interpretation must be confirmed by more detailed onshore analysis and results of planned drilling of sediments on the incoming plate. The modern trench wedge is composed of channel-filling deposits that are currently underthrusting beneath the accretionary prism. The plate boundary frontal thrust is located at a stratigraphic horizon of late Miocene mud, which appears to be similar to the Shikoku Basin facies that hosts the décollement in the Muroto and Ashizuri transects (Moore, Taira, Klaus, et al., 2001).

The sediments of lithologic Unit I at Site C0006 are interpreted to represent a transition from trench to slope deposition; thus, the boundary between Units I and II records the uplift of trench material into the prism, and its age potentially provides a constraint on the timing of frontal thrust activity (Fig. F4). Sediments of Unit I are younger or the same age as sediments filling the basin behind the thrust. The age of the boundary is ~0.9 Ma. Taking into account the relative plate motion velocity between the overriding Japanese islands and the Philippine Sea plate, ~4 cm/y (Seno et al., 1993), the relative slip distance has to be ~40 km. No frontal accretion during this period means that the plate boundary frontal thrust has large displacement (total relative motion minus horizontal shortening of the accretionary prism). How such a large amount of slip has concentrated within the fault zone and how the evolution of the fault zone has affected the characteristic features in this region is one of the important issues of postcruise research.

The accretionary prism in the frontal thrust region is deformed by thrusting, as visible on the seismic profiles (Moore et al., 2007) and LWD data (Kinoshita et al., 2008). Most of the thrusts inferred from the seismic profile and LWD were confirmed from age reversals, fault zones sampled in cores, and repetition of specific strata, but some additional faults were also defined during Expedition 316 drilling. Chemical analysis of interstitial fluid and microbial habitat around most of these "intraprism" thrusts do not indicate any signal of active fluid flow.

In contrast to such intraprism thrusts, many normal faults are developed at the core scale and appear to be the youngest deformation feature. Clear slope-parallel mass sliding is observed from the submarine topography, seismic profiles, and shallow cores. These facts suggest that the taper angle of the prism is presently above the critical wedge taper angle and is unstable; there is evidence that the system is currently in a period of collapse. These preliminary interpretations of the frontal part of the accretionary prism have to be investigated from the physical and hydrological properties of accretionary prism material in postcruise research. Extremely low heat flow observed at Sites C0006 and C0007 might be related to stratigraphic or structural fluid pathways developed in this region. Porosity data suggest that a considerable thickness of material has been removed from the surface at Sites C0006 and C0007. Porosities are quite low at shallow depths below the surface, reaching 48% at 5 and 34 m CSF at Sites C0006 and C0007, respectively. In contrast, porosities do not decrease to <50% until ~150 or ~200 m CSF, respectively, at prism Site C0004 and slope basin Site C0008.

Frontal thrust fault rocks were successfully recovered and preliminary visual core description shows shear localization and repeated slip in the fault zone; relationships between apparent repeated localized slip and coseismic and/or interseismic slip are a primary component of postcruise science research. Shipboard geochemical analyses suggest possible fluid flow around the frontal thrust. Episodicity of fluid flow and relationships between pore pressure and slip along the thrust are keys to understanding the dynamics of this part of accretionary prism and plate boundary thrust.