Search

The JFAST drilling project endeavored to establish the stress state on the shallow subduction megathrust that slipped during the M9 Tohoku earthquake. Borehole breakout data from the drillhole can constrain both the orientation and magnitude of the principal stresses. Here we reanalyze that data to refine our understanding of the stress state on the fault. In particular, we: (1) Improve the identification of breakouts, (2) Consider a fuller range of stress states consistent with the data, and (3) Incorporate new and more robust laboratory constraints on rock strength. The original conclusion that the region is in a normal faulting regime after the earthquake is strengthened by the new analysis. The combined analysis suggests the earthquake released sufficient elastic strain energy to reset the local stress field.

IODP Expedition 314 acquired annular pressure while drilling measurements in six holes extending from the outer forearc basin to the base of the trench slope. These data provide pressures inside the borehole near the bit during drilling; the pressures reflect the cuttings load in the borehole, and the viscous resistance to flow of fluids up the borehole due to pumping from the rig floor and due to fluid exchange with the formation. The forearc basin site provides a reference locality that extends to 1402 m and whose pressure curve is apparently controlled dominantly by the cuttings load and the rig-floor pumping rate. All other holes show at least initial pressure curves that are very similar to that at the forearc basin site, despite pumping rates varying by about 20–30%. Some small variations in pressure curves can be positively correlated to the penetration rate that controls cutting load. Two holes crossing a fracture zone show pressure anomalies extending to above-lithostatic conditions, and modeled fluid influx rates of 3000 l/m or 150% of the rig-floor pumping rate. A third hole, probably in the same highly overpressured zone, shows a near lithostatic pressure value just prior to losing the bottom hole assembly due to stuck pipe. The extraordinarily high pressures occur in the upper portion of the hanging wall of the megasplay fault zone and may reflect residual conditions from the 1944 M 8.1 earthquake. A penetration of the lower boundary of the megasplay fault zone shows only minor fluid pressure anomalies around the fault surface. A hole through the frontal thrust records fluid pressure anomalies in sandy sediments of the footwall, apparently due to expulsion of fluids from loading by the hanging wall.

Active faults slip at different rates over the course of the seismic cycle: earthquake slip ( c . 1 m s −1 ), interseismic creep ( c . 10–100 mm year −1 ) and intermediate rate transients (e.g. afterslip and slow slip events). Studies of exhumed faults are sometimes able to identify seismic slip surfaces by the presence of frictional melts, and slow creep by textures diagnostic of rate-limited plastic processes. The Pasagshak Point Thrust preserves three distinct fault rock textures, which are mutually cross-cutting, and can be correlated to different strain rates. Ultrafine-grained black fault rocks, including pseudotachylyte, were formed during seismic slip on layers up to 30 cm thick. Well-organized S – C cataclasites 7–31 m thick were formed by slow creep, with pressure solution as a dominant, rate-limiting mechanism. These must have formed at strain rates consistent with long-term plate-boundary motion, but solution-creep healing acted to reduce porosity of the cataclasites and eventually restricted fluid connectivity such that creep by this mechanism could not continue. Disorganized, non-foliated, rounded clast cataclasites were formed at shear rates faster than solution creep and are interpreted as representing shear at intermediate strain rates. These could have formed during afterslip or delocalization of slip associated with an earthquake rupture.

Active faults slip at different rates over the course of the seismic cycle: earthquake slip (c. 1 m s−1), interseismic creep (c. 10–100 mm year−1) and intermediate rate transients (e.g. afterslip and slow slip events). Studies of exhumed faults are sometimes able to identify seismic slip surfaces by the presence of frictional melts, and slow creep by textures diagnostic of rate-limited plastic processes. The Pasagshak Point Thrust preserves three distinct fault rock textures, which are mutually cross-cutting, and can be correlated to different strain rates. Ultrafine-grained black fault rocks, including pseudotachylyte, were formed during seismic slip on layers up to 30 cm thick. Well-organized S–C cataclasites 7–31 m thick were formed by slow creep, with pressure solution as a dominant, rate-limiting mechanism. These must have formed at strain rates consistent with long-term plate-boundary motion, but solution-creep healing acted to reduce porosity of the cataclasites and eventually restricted fluid connectivity such that creep by this mechanism could not continue. Disorganized, non-foliated, rounded clast cataclasites were formed at shear rates faster than solution creep and are interpreted as representing shear at intermediate strain rates. These could have formed during afterslip or delocalization of slip associated with an earthquake rupture.

International audience; To better understand the distribution of three dimensional stress states in the Nankai subduction zone, southwest Japan, we review various stress-related investigations carried out in the first and second stage expeditions of the Nankai Trough Seismogenic Zone Experiment (NanTroSEIZE) by the Integrated Ocean Drilling Program (IODP) and compile the stress data. Overall, the maximum principal stress σ1 in the shallower levels (~ 1 km below seafloor or in underlying accretionary prism) with σ1 becoming horizontal is also suggested at all deeper drilling sites. We also make a comparison of the stress state in the hanging wall of the frontal plate-interface between Site C0006 in the Nankai and Site C0019 in the Japan Trench subduction zone drilled after the 2011 Mw 9.0 Tohoku-Oki earthquake. In the Japan Trench, a comparison between stress state before and after the 2011 mega-earthquake shows that the stress changed from compression before the earthquake to extension after the earthquake. As a result of the comparison between the Nankai Trough and Japan Trench, a similar current stress state with trench parallel extension was recognized at both C0006 and C0019 sites. Hypothetically, this may indicate that in Nankai Trough it is still in an early stage of the interseismic cycle of a great earthquake which occurs on the décollement and propagates to the toe (around site C0006).

Electrical images recorded with Resistivity-At-Bit (RAB) from two sites drilled during Ocean Drilling Program (ODP) Leg 196 were analyzed to study the effects of subduction at the Nankai margin. For the first time in the history of scientific deep-sea drilling in ODP, in situ complete borehole images of the decollement zone were obtained. Analyses of all drilling-induced fracture data indicated that the maximum horizontal compressive stress (SHmax) axes have an azimuth of 303°, and analyses of breakout data from RAB images indicated an azimuth of 310°. These azimuths approximate the convergence direction of the Philippine Sea plate towards the Eurasian plate. The frontal thrust at Site 808 was encountered at about 389 mbsf. Density, porosity, resistivity, and gamma ray data change across the frontal thrust. The decollement zone at the deformation front was identified between 937 and 965 mbsf. The base of the decollement is sharply defined as the maximum extent of conductive fracturing and is marked by abrupt changes in physical properties [Mikada, H., Becker, K., Moore, J.C., Klaus, A., Austin, G.L., Bangs, N.L., Bourlange, S., Broilliard, J., Bruckmann, W., Corn, E.R., Davis, E.E., Flemings, P.B., Goldberg, D.B., Gulick, S.S., Hansen, M.B., Hayward, N., Hills, D.J., Hunze, S., Ienaga, M., Ishiguro, H., Kinoshita, M., Macdonald, R.D., McNeill, L., Obana, S., Hong, O.S., Peacock, S., Pettigrew, T.L., Saito, S., Sawa, T., Thaiprasert, N., Tobin, H.J., Tsurumi, H., 2002. Proc. ODP, Initial Rep., 196, College Station, TX, (Ocean Drilling Program)]. The upper boundary of the decollement is marked by several sets of conductive fractures and by high variability in physical properties. The decollement zone is characterized by intense brittle fracturing. These fractures are considered to be the consequence of cyclic stresses and high fluid pressures in this zone. We analyzed fracture dips and their orientations at both sites and found that they are all consistent with a unique stress field model surrounding the two sites.

In central California, Maastrichtian–Danian shales of the Moreno Formation preserve a fluid migration system that developed along the western margin of the former Great Valley forearc basin. The system consists of a network of interconnected sandstone intrusions linked to overlying fossiliferous carbonates whose geochemistry, fauna, and petrology are characteristic of active cold seeps. The system is approximately 800 m thick and represents episodic migration and seafloor expulsion of fluids over at least 0.5×106 years. This locality has the most extensive exposure yet discovered of a complete seep system, from underlying fluid pathways to seep deposits and associated communities.

Fracture porosity in the decollement zone of Nankai accretionary wedge is estimated by comparison of porosity measured on cores during Ocean Drilling Program Leg 131 and porosity calculated from resistivity logs acquired during Leg 196 using Logging While Drilling. Resistivity is converted to formation factor considering both pore fluid conductivity and surface conductivity of clay particles. Pore fluid conductivity is calculated from temperature and ion concentration in interstitial water, whereas surface conductivity is calculated from cationic exchange capacity data and exchangeable cation concentrations. Finally the formation factor is converted to porosity using the generalized Archie’s law. The decollement appears as a zone of compacted rock where dilatant fractures have developed. The contrast between resistivity–porosity and core porosity is used to estimate fracture porosity in the decollement, assuming that the total conductivity is the result of fracture network and rock fragment conductivities, behaving as resistors in parallel, in the direction of the fracture network. Fracture porosity increases downward in the decollement zone from 1.8% to 8.5%. This suggests pore pressure in the decollement zone is higher than the pore pressure estimated from compaction curves (excess pore pressure ratio of 0.47). A possible explanation is that dilatancy is associated with a high pressure transient. The migration of a pressure wave along the decollement could occur at a velocity of 500 m/yr if the permeability of the dilated zone is higher than 10−12 m2. The characteristic time for transient dissipation by diffusion in the footwall and hanging wall of the decollement is estimated to be 100–1000 years. Coexistence of dilatant and compactive shear localization structures is observed within the wedge and in the main fault zones. However, only the decollement is currently dilated by fluids. We propose that fluids are injected into the decollement zone during or after fracturing and that initial shear localization is always compactive and occurs ahead of the fluid injections. This sequence of events could occur during each fluid migration and slip event, constituting an increment of decollement propagation.

International audience; We have used sandbox experiments to explore the mechanics of the frontal prism structures documented by seismic reflection data and new borehole from IODP Expedition 343 (JFAST). This study investigated the effects of down-dip (normal to trench axis) variations in frictional resistance along a decollement on the structural development of the frontal wedges near subduction zones. Interpretation of seismic reflection images indicates that the wedge has been effected by trench-parallel horst-and-graben structures in the subducting plate. We performed sandbox experiments with down-dip patches of relatively high and low friction on the basal decollement to simulate the effect of variable coupling over subducting oceanic plate topography. Our experiments verify that high frictional resistance on the basal fault can produce the internal deformation and fault-and-fold structures observed in the frontal wedge by the JFAST expedition. Subduction of patches of varying friction caused a temporal change in the style of internal deformation within the wedge and gave rise to two distinctive structural domains, separated by a break in the surface slope of the wedge: (i) complexly deformed inner wedge with steep surface slope and (ii) shallow taper outer wedge, with a sequence of imbricate thrusts. Our experiments further demonstrate that the topographic slope-break in the wedge develops when the hinterland part of the wedge essentially stops deforming internally, leading to in-sequence thrusting with the formation of an outer wedge with low taper angle. For a series of alternate high and low frictional conditions on the basal fault the slope of the wedge varies temporally between a topographic slope-break and uniformly sloping wedge.

The 2011 Mw9.0 Tohoku-oki earthquake ruptured to the trench with maximum coseismic slip located on the shallow portion of the plate boundary fault. To investigate the conditions and physical processes that promoted slip to the trench, Integrated Ocean Drilling Program Expedition 343/343T sailed 1 year after the earthquake and drilled into the plate boundary ∼7 km landward of the trench, in the region of maximum slip. Core analyses show that the plate boundary decollement is localized onto an interval of smectite-rich, pelagic clay. Subsidiary structures are present in both the upper and lower plates, which define a fault zone ∼5–15m thick. Fault rocks recovered from within the clay-rich interval contain a pervasive scaly fabric defined by anastomosing, polished, and lineated surfaces with two predominant orientations. The scaly fabric is crosscut in several places by discrete contacts across which the scaly fabric is truncated and rotated, or different rocks are juxtaposed. These contacts are inferred to be faults. The plate boundary decollement therefore contains structures resulting from both distributed and localized deformation. We infer that the formation of both of these types of structures is controlled by the frictional properties of the clay: the distributed scaly fabric formed at low strain rates associated with velocity-strengthening frictional behavior, and the localized faults formed at high strain rates characterized by velocity-weakening behavior. The presence of multiple discrete faults resulting from seismic slip within the decollement suggests that rupture to the trench may be characteristic of this margin.

The application of three-dimensional seismic reflection and coherence imaging to the study of the decollement zone of the Barbados Ridge accretionary complex has provided new insights into the relationships among internal structure, fluid flow, and previously unrecognized strike-slip faulting. Combined coherence and seismic amplitude imaging of the decollement zone reveal anomalous northeast-trending lineaments parallel to and abutting zones of high- amplitude, negative-polarity reflections. Analysis of these lineaments shows them to be penetrative structures dipping southeast with apparent reverse dip-slip offset. Isopach mapping of the accretionary wedge indicates significant right-lateral displacement across these structures. These faults apparently channel fluid flow within the decollement zone, and the prominent northeast- trending conduits so formed are readily visible as high-amplitude, negative-polarity reflections. Additionally, north-northeast– trending zones of variable coherence and high positive amplitude are inferred barriers to up-structure fluid-migration pathways. Movement along strike-slip structures probably alternates with displacement along the decollement zone. Northeast- trending strike-slip faults extend for >13 km, crossing the length of the survey area and into the incoming oceanic plate. Active arc-oblique strike-slip faulting of the decollement zone beneath the Barbados Ridge accretionary wedge implies a stress regime in that σ 1 is fixed and σ 2 and σ 3 either transpose with time or are nearly equal. This state of stress may be a common occurrence in forearc tectonism and may have led to the formation of many, as yet unrecognized, arc-oblique strike-slip faults at convergent margins.

A volume of three-dimensional seismic reflection data, acquired in 1992, imaged the decollement beneath the northern Barbados Ridge accretionary prism revealing reflection amplitude and waveform variations attributed to fluid accumulations along the plate boundary fault. We model the seismic reflection by inversion for seismic impedance (the product of velocity and density) throughout the 5 x 25 km survey area and thus map physical property variations. In 1997, Ocean Drilling Program Leg 171A penetrated the protodecollement and decollement at five sites with a logging-while-drilling (LWD) tool to log density and other physical properties of the decollement. We construct a regional map of density, and inferred porosity, within the decollement from seismic models calibrated with LWD density data. In the sediments out in front of the trench the protodecollement forms in a radiolarian-rich Miocene mudstone with an anomalously high porosity (70-75%) that appears as a pervasive, inherent characteristic of this interval seaward of the deformation front. In the decollement beneath the wedge a consolidation trend of decreasing porosity runs perpendicular to the deformation front with porosity decreasing from 70% at the wedge toe to 50% 4 km from the wedge toe. A second, distinct trend also forms along a 10-km-long, 1- to 2-km-wide, NE-SW zone in which porosity is 70%, as high as it is in the protodecollement. This zone can be explained as an area of the decollement where fluid accumulations develop by maintaining high fluid content. We postulate that high fluid content is maintained by continuous recharge flowing into and along this channel. This porosity distribution within the decollement also strongly influences fluid migration into the overlying accretionary wedge and is directly associated with fluid charging of ramps and out-of-sequence thrusts above the decollement.

Deep Drilling for Earthquake Clues The 2011 M w 9.0 Tohoku-Oki earthquake and tsunami were remarkable in many regards, including the rupturing of shallow trench sediments with huge associated slip (see the Perspective by Wang and Kinoshita ). The Japan Trench Fast Drilling Project rapid response drilling expedition sought to sample and monitor the fault zone directly through a series of boreholes. Chester et al. (p. 1208 ) describe the structure and composition of the thin fault zone, which is predominately comprised of weak clay-rich sediments. Using these same fault-zone materials, Ujiie et al. (p. 1211 ) performed high-velocity frictional experiments to determine the physical controls on the large slip that occurred during the earthquake. Finally, Fulton et al. (p. 1214 ) measured in situ temperature anomalies across the fault zone for 9 months, establishing a baseline for frictional resistance and stress during and following the earthquake.

A multi-channel seismic survey (MCS) yields high resolution images of geologic structures within the Oregon accretionary prism. In addition to revealing the internal geometry of the prism a prominent bottom simulating reflector (BSR) provides indirect estimates of temperature gradients. The BSR marks the phase transition, within the pore space of the sediments, of small amounts of solid gas hydrate above the reflector to free gas below. Although the absolute value of temperature gradients depends on assumptions of gas hydrate stability conditions, velocity functions, and seafloor temperature, relative differences reflecting regional trends and local anomalies are unaffected by the assumptions involved in calculation of temperature gradient from the position of the BSR. Broad-scale landward decreases in temperature gradient reflect the increasing depth to the oceanic crust across the accretionary prism. Correlation of anomalously high temperature gradients with two out-of-sequence thrusts indicates that these structures are currently acting as conduits for warm fluids.

The thickness of an active plate boundary fault is an important parameter for understanding the strength and spatial heterogeneity of fault behavior. We have compiled direct measurements of the thickness of subduction thrust faults from active and ancient examples observed by ocean drilling and fi eld studies in accretionary wedges. We describe a general geometric model for subduction thrust decollements, which includes multiple simultaneously active, anastomosing fault strands tens of meters thick. The total thickness encompassing all simultaneously active strands increases to ~100–350 m at ~1–2 km below seafl oor, and this thickness is maintained down to a depth of ~15 km. Thin sharp faults representing earthquake slip surfaces or other discrete slip events are found within and along the edges of the tens-ofmeters- thick fault strands. Although fl attening, primary inherited chaotic fabrics, and fault migration through subducting sediments or the frontal prism may build melange sections that are much thicker (to several kilometers), this thickness does not describe the active fault at any depth. These observations suggest that models should treat the subduction thrust plate boundary fault as

Stressed Out Large seismic events such as the 2011 magnitude 9.0 Tohoku-Oki earthquake can have profound effects not just on the severity of ground motion and tsunami generation, but also on the overall state of the crust in the surrounding regions. Lin et al. (p. 687 ) analyzed the stress 1 year after the Tohoku-Oki earthquake and compared it with the estimated stress state before the earthquake. In situ resistivity images were analyzed from three boreholes drilled into the crust across the plate interface where the earthquake occurred. Stress values indicate a nearly complete drop in stress following the earthquake such that the type of faulting above the plate boundary has changed substantially. These findings are consistent with observations that the sea floor moved nearly 50 meters during the earthquake.

Either thrusting of higher over lower impedance sediment or reduction of impedance locally in the fault zone can produce negative-polarity reflections along the protothrusts and the frontal thrust of the Oregon accretionary prism. The vertical displacement of protothrusts averages only 27 m, is uncorrelated to the occurrence of polarity reversals, and is insufficient to generate the impedance contrasts associated with the negative-polarity reflections. The negative-polarity reflections from the protothrusts are probably due to localized impedance reduction in the fault zone. Waveform modeling of these negative-polarity reflections indicates they develop due to a single interface with negative impedance contrast. This seismic signal could be produced by a fault zone with a sharp upper boundary with low impedance fault material and a transition down section to sediment of higher impedance. An impedance inversion due to the 1-km throw on the frontal thrust could create the observed negative-polarity reflections. Interval velocities do not decrease across the frontal thrust, but core-scale velocity measurements indicate a velocity decrease in the frontal thrust. Therefore the negative-polarity reflections are apparently produced by a localized decrease in impedance. Waveform modeling of frontal thrust reflections suggests the low impedance interval is about 20 m thick with discrete upper and lower boundaries. Moderately dipping faults are an efficient path for fluid expulsion across the turbidites with low equivalent vertical permeability. Fluid migration from the over-pressured section up the faults can locally reduce the velocity sufficiently to cause the negative-polarity reflections.

[1] Annular pressure while drilling data shows high fluid overpressures at Site C0001 in part of the megasplay fault zone of the NanTroSEIZE transect across the subduction zone of SW Japan. Mostly standard annular pressures while drilling occur at three other sites, including two penetrating major faults. The two holes at Site C0001 show a step up to lithostatic annular pressure at about 500 mbsf, following initial indicators of overpressure at about 375 mbsf (meters below seafloor). The annular pressure remains high and increasing to total depth of 1000 mbsf. Seismic lines through the site show bright reflectors in the zone of initial annular pressure increase. Borehole images, sonic velocities, and resistivity all suggest a zone of fractures, from about 490 to 630 mbsf. A hydraulic model of the fluid system explains the observed pressures by influx of formation fluid at about 500 mbsf. The combination of a natural influx of 3300 l/m plus 2200 l/m from the drilling system can explain the observed annular pressures. The highly fractured zone that bleeds fluids to the borehole may be sealed by a localized zone of compressive stress or by overlying gas hydrates.

[1] Four drill sites of IODP NanTroSEIZE Stage 1 Expedition transected the Nankai Trough, offshore SW Japan, from the deformation front to the Kumano fore-arc basin. Borehole resistivity images from the logging-while-drilling (LWD) data were analyzed to extract orientations of faults, fractures, and bedding planes to examine the structural styles. On the basis of these features, drilling intervals were classified into fore-arc basin deposits, surface slope sediments, and deformed accretionary wedge, and these can be compared with characteristics from seismic profiles and core structural data. Bedding orientations identified in these three data sets are generally comparable, but the difference in resolution between the data sets produces different results in interpretation where geology is highly deformed or includes finer internal structures. Faults can also be correlated between these three data sets, but the differences in their appearance require special attention for accurate correlation. Many faults imaged in seismic profiles actually consist of microfracture systems, as shown in cores, that can also be identified in borehole images. Some clear faults in seismic profiles cannot be identified in borehole images, probably because of their minimal resistivity contrast with the surrounding rocks or a more complex fault zone at the borehole scale. These results suggest that these three data sets can be used to extract not only the general structure but also different styles of deformation at different scales from core samples (mm to cm), to LWD (mm to 10 m), to seismic (10 m to tens of km). This correlation requires a deep understanding of the resolution and shortcomings of each methodology.

Resistivity at the bit tools typically provide images of wellbore breakouts only a few minutes after the hole is drilled. In certain cases images are taken tens of minutes to days after drilling of the borehole. The sonic caliper can also image borehole geometry. We present four examples comparing imaging a few minutes after drilling to imaging from about 30 min to 3 days after drilling. In all cases the borehole breakouts widen with time. The tendency to widen with time is most pronounced within a few hundred meters below the seafloor (mbsf), but may occur at depths greater than 600 mbsf. In one example the widening may be due to reduced borehole fluid pressure that would enhance borehole failure. In the three other cases, significant decreases in fluid pressure during temporal evolution of breakouts are unlikely. The latter examples may be explained by time-dependent failure of porous sediments that are in an overconsolidated state due to drilling of the borehole. This time-dependent failure could be a consequence of dilational deformation, decrease of pore fluid pressure, and maintenance of sediment strength until migrating pore fluids weaken shear surfaces and allow spallation into the borehole. Breakout orientations, and thus estimates of stress orientations, remain consistent during widening in all four cases. In vertical boreholes, breakouts wider than those initially estimated by resistivity imaging would result in higher estimates of horizontal stress magnitudes. Because the vertical overburden stress is fixed, higher estimated horizontal stresses would favor strike-slip or thrust faulting over normal faulting.

The Hoh rock assemblage of the western Olympic Peninsula, Washington, was Deposited in Eocene to Miocene time, and accreted, deformed, uplifted, and eroded by the late Miocene. Northeast-dipping fault zones within the Hoh are characterized by intensely deformed mud matrix melange, and they range from ∼200 m to >1,000 m thick; kinematic indicators within the matrix indicate thrusting to the west-southwest. Both hanging-wall and footwall sections are strongly folded and faulted. Wall-rock sections are lithostratigraphically distinct, with hanging walls composed of massive coarse sandstone and conglomerate, and footwall sections composed of thin- to medium- bedded sandstone and siltstone. Folds within the wall-rock sections are oblique to, and are truncated by, the later fault-zone melanges, indicating that the fault zones postdate folding. Evidence of fluid involvement in the Hoh rocks includes petroliferous outcrops, carbonate and clay cement, and veins of carbonate and laumontite. The progressive evolution of veins and cements (paragenetic sequence) can be interpreted in terms of crystallization under conditions of increasing consolidation, decreasing salinity, and changing fluid composition. Suppressed carbonate luminescence may reflect crystallization under reducing conditions, perhaps related to the sulfate reduction zone, the presence of hydrocarbons, or clay dehydration. Although deformational features may be associated with the earliest phase of veining, most of the localized grain breakage and matrix shear zones formed after the early micritic veins and before the last phase of sparitic carbonate (+laumontite) veins. Laumontite occurred syn-deformationally and in equilibrium with carbonate, and also post-deformationally. The paragenetic sequences observed in all samples from the Hoh rock assemblage can be correlated, suggesting that the sources and processes controlling the fluid emplacement were similar across a wide region. In addition, the paragenetic sequence in the Hoh rocks is similar to that documented in autochthonous coeval basins of southwest Washington. This similarity suggests that the precipitates, formed during the evolution of the Hoh rocks, involved fluids that originated within the matrix and did not involve exotic or far-traveled fluids.

Sediment offscraping and accretion takes place predominantly by landward-vergent thrusting along most of the Oregon accretionary prism north of latitude 44°50'N. This indicates very low basal shear stress, probably due to near-lithostatic fluid pressure on the decollement horizon. Recent multichannel seismic and sidescan sonar studies have demonstrated the existence of an oblique, basement-involved, left-lateral strike-slip fault, termed the Wecoma fault. It cuts abyssal plain sediments and oceanic basement of the subducting plate. Where the Wecoma fault intersects the wedge thrust front at 45°09'N, accretion to the wedge occurs through an anomalous seaward-vergent thrust. Linear, fault-parallel erosional gullies, which trend obliquely across the slope on the seaward flank of the frontal anticline in the accretionary wedge, are interpreted as marking the surface traces of splays of the Wecoma fault. Data collected during six Alvin dives on and around these gullies demonstrate that fluid expulsion from the deforming sediments is preferentially concentrated along the fault traces. Evidence includes the presence of chemosynthetic biological communities, pervasive veining and fracturing of the rock, and extensive carbonate cementation, all forming preferentially within the gullies. This strike-slip fault apparently acts as a high-angle fluid escape conduit, channeling fluids from deep within the wedge to the surface. We propose that this fluid conduit causes the local inversion of thrusting to seaward vergence by allowing the escape of highly overpressured fluids from the incipient decollement horizon. The consequent increase in basal effective normal stress favors the development of a seaward-vergent thrust. The discovery of this strike-slip fault acting as an efficient fluid conduit to the surface of the wedge emphasizes the generally unappreciated importance of strike-slip (and normal) faults in dewatering accretionary prisms. Such dewatering features may significantly influence the development of large-scale structures, and they have broad applicability to other tectonic settings, including collision zones and continental fold-and-thrust belts.

We constrain the orientations and magnitudes of in situ stress tensors using borehole wall failures (borehole breakouts and drilling-induced tensile fractures) detected in four vertical boreholes (C0002, C0001, C0004, and C0006 from NW to SE) drilled in the Nankai accretionary wedge. The directions of the maximum horizontal principal stress (SHmax), indicated by the azimuths of borehole wall failures, are consistent in individual holes, but those in C0002 (margin-parallel SHmax) are nearly perpendicular to those in all other holes (margin-normal SHmax). Constrained stress magnitudes in C0001 and C0002, using logged breakout widths combined with empirical rock strength derived from sonic velocity, as well as the presence of the drilling-induced tensile fractures, suggest that the stress state in the shallow portion of the wedge (fore-arc basin and slope sediment formations) is predominantly in favor of normal faulting and that the stress state in the deeper accretionary prism is in favor of probable strike-slip faulting or possible reverse faulting. Thus, the stress regime appears to be divided with depth by the major geological boundaries such as unconformities or thrust faults. The margin-perpendicular tectonic stress components in the two adjacent sites, C0001 and C0002, are different, suggesting that tectonic force driven by the plate pushing of the Philippine Sea plate does not uniformly propagate. Rather, the stress field is inferred to be influenced by additional factors such as local deformation caused by gravitation-driven extension in the fore arc and thrusting and bending within individual geologic domains.

[1] A 1.6 km riser borehole was drilled at site C0009 of the NanTroSEIZE, in the center of the Kumano forearc basin, as a landward extension of previous drilling in the southwest Japan Nankai subduction zone. We determined principal horizontal stress orientations from analyses of borehole breakouts and drilling-induced tensile fractures by using wireline logging formation microresistivity images and caliper data. The maximum horizontal stress orientation at C0009 is approximately parallel to the convergence vector between the Philippine Sea plate and Japan, showing a slight difference with the stress orientation which is perpendicular to the plate boundary at previous NanTroSEIZE sites C0001, C0004 and C0006 but orthogonal to the stress orientation at site C0002, which is also in the Kumano forearc basin. These data show that horizontal stress orientations are not uniform in the forearc basin within the surveyed depth range and suggest that oblique plate motion is being partitioned into strike-slip and thrusting. In addition, the stress orientations at site C0009 rotate clockwise from basin sediments into the underlying accretionary prism.

The central Oregon margin spans a regional transition in accretionary structures from seaward-verging in the south to landward-verging in the north. New multichannel seismic (MCS) data image both landward- and seaward-vergent provinces along the central and northern Oregon margin. Landward-vergence is characterized by a deep decollement, with deformation distributed across a broad lower continental slope in a coherent structural style. In the landward-vergent area, virtually all 4 km of incoming trench sediments overthrust the preceding thrust sheet, forming fault-bend folds and a distinctive ridge/trough morphology. This style of landward-vergence is not explained by existing models. In contrast, seaward-vergence correlates with a shallower decollement, approximately 1.4 km above the oceanic crust, and a more intense style of deformation within a narrower slope. Initial thickening of the trench sediments occurs across a well-developed protothrust zone. The frontal thrust forms a ramp-anticline that is cut by a prominent backthrust. Previously observed seafloor vent sites in both regions correlate with thrusts that exhibit high-amplitude, reversed-polarity reflections suggestive of enhanced porosity along the faults. Potential fluid sources and migration paths are strongly influenced by changes in the level of the decollement and vergence along the margin. Abrupt changes in structural style occur both along strike and updip, and are bounded by two sets of oblique-slip faults. Three NW-striking left-lateral faults are imaged in both MCS and SeaBeam data. Plunging anticlines developed along the NW-striking faults are venting fluids and were previously interpreted as mud volcanoes. The deformation front is locally disrupted where these faults intersect the prism, but they appear to have limited influence on the structural evolution of the prism. In contrast, the NE-striking right-lateral faults are confined to the deforming sediments of the upper plate. These faults interact with the thrusts within the prism, forming a rhomboidal pattern of three-dimensional deformation.

[1] The Uganik Thrust is a fossil out-of-sequence thrust fault which was active over a period of ∼3 Ma during the early Tertiary until activity ceased with the subduction of the Kula-Farallon spreading ridge at ∼57 Ma. During this period the fault experienced at least ∼1 km of throw and developed a strongly asymmetric damage zone. The brittle damage zone in the footwall of the fault acted as a conduit for fluid advection during the active faulting. A similar asymmetrical footwall damage zone has been interpreted as a fluid conduit at the Nobeoka Thrust, Shimanto Belt, SW Japan. Thermal indicators in the uppermost footwall give similar maximum paleotemperatures to those in the hanging wall (∼280°C), while previous work elsewhere in the footwall formation suggests maximum burial temperatures of ∼240°C. In this case, similar to the Irish Canyon thrust in the Franciscan accretionary complex, the location of the thermal anomaly is spatially offset from the structural fault which caused it owing to thermal overprinting in the vicinity of the fault.

Fault zones focus fluid expulsion in the muddy northern Barbados Ridge accretionary prism with fault-parallel permeabilities about 1000 times greater than intergranular permeabilities in the adjacent sediment. In the Oregon prism the low bedding-perpendicular permeability (due to mudstones) inhibits intergranular dewatering; however, intergranular flow is concentrated where submarine erosion breaches high permeability sandy layers. Even so, faults can capture fluid flow from these exposed sandy layers suggesting the faults have a still higher permeability. Such observations coupled with laboratory measurements permeabilities suggest that faults off Oregon may have fault-parallel permeabilities at least 10-10000 times greater than the adjacent sediments. Results from Barbados and Oregon suggest fluid flow is concentrated along the most active faults. At the toe of prisms the fault zones are being progressively loaded by the thickening wedge and are undergoing compaction. Preliminary experiments show that permeability decreases relative to the surrounding wall rocks along faults within this compactive deformation regime; we believe that these faults must undergo dilation, perhaps linked to transient increases in pore pressure if they are to be preferential fluid conduits. Farther upslope erosion exposes rocks that are more consolidated, commonly more cemented, and generally of lower intergranular permeability than rocks of equivalent burial further seaward. Because of their lithification and overconsolidation these rocks dilate during faulting, locally enhancing fracture permeability. In such dilative regimes, faults become evermore focused zones of fluid expulsion relative to occluded intergranular pathways.

Earthquakes occur along the plate-boundary thrusts underlying accretionary prisms and along out-of-sequence thrusts that cut through prisms. Thermal models suggest that the earthquakes on the plate-boundary thrusts initiate in a temperature range of 125 o C to about 350 o C. Because syndeformational diagenetic and metamorphic alterations recorded in accretionary prisms have specific temperature ranges, the alterations and the associated deformation can be correlated to the temperature range that accretionary prisms are seismogenic. Comparison of accreted rocks deformed above, within, and below the seismogenic zone suggests characteristics of rocks at seismogenic depths that may make them earthquake-prone.

Preface Part I. Introduction 1. The Seismogenic Zone of Subduction Thrust Faults: Introduction, by Timothy H. Dixon and J. Casey Moore 2. The Seismogenic Zone of Subduction Thrust Faults: What We Know and Don't Know, R. D. Hyndman Part II. The Incoming Plate 3. Sediment Inputs to Subduction Zones: Why Lithostratigraphy and Clay Mineralogy Matter, by Michael B. Underwood 4. The Thermal State of 18 -24 Ma Upper Lithosphere Subducting Below the Nicoya Peninsula, Northern Costa Rica Margin, by M. Hutnak, A. T. Fisher, C. A. Stein, R. Harris, K. Wang, E. Silver, G. Spinelli, M. Pfender, H. Villinger, R. MacKnight, P. Costa Pisani, H. DeShon, and C. Diamente 5. Influence of Subducting Topography on Earthquake Rupture, by Susan L. Bilek Part III. Convergent Margin Structure, Fluids and Subduction Thrust Evolution 6. Pore Pressure and Fluid Flow in the Northern Barbados Accretionary Complex: A Synthesis, by Barbara A. Bekins and Elizabeth J. Screaton 7. Pore Pressure within Underthrust Sediment in Subduction Zones, by Demian M. Saffer 8. Deformation and Mechanical Strength of Sediments at the Nankai Subduction Zone: Implications for Prism Evolution and Decollement Initiation and Propagation, by Julia K. Morgan, Elizabeth B. Sunderland, ne E. Blanche Ramsey, and Maria V. S. Ask 9. The Nicaragua Convergent Margin: Seismic Reflection Imaging of the Source of a Tsunami Earthquake, by Kirk D. McIntosh, Eli A. Silver, Imtiaz Ahmed, Arnim Berhorst, Cesar R. Ranero, Robyn K. Kelly, and Ernst R. Flueh 10. How Accretionary Prisms Elucidate Seismogenesis in Subduction Zones, by J. Casey Moore, Christie Rowe, and Francesca Meneghini Part IV. Laboratory Studies 11. Friction of the Smectite Clay Montmorillonite: A Review and Interpretation of Data, by Diane E. Moore and David A. Lockner 12. Fault Friction and the Upper Transition from Seismic to Aseismic Faulting, by Chris Marone and Demian M. Saffer 13. Laboratory-Observed Faulting in Intrinsically and Apparently Weak Materials: Strength, Seismic Coupling, Dilatancy, and Pore-Fluid Pressure, by N. M. Beeler Part V. Seismic and Geodetic Studies 14. Asperities and Quasi-Static Slips on the Subducting Plate Boundary East of Tohoku, Northeast Japan, by Akira Hasegawa, Naoki Uchida, Toshihiro Igarashi, Toru Matsuzawa, Tomomi Okada, Satoshi Miura, and Yoko Suwa 15. Anomalous Earthquake Ruptures at Shallow Depths on Subduction Zone Megathrusts, by Thorne Lay and Susan Bilek 16. Secular, Transient and Seasonal Crustal Movements in Japan from a Dense GPS Array: Implication for Plate Dynamics in Convergent Boundaries, by Kosuke Heki 17. Elastic and Viscoelastic Models of Crustal Deformation in Subduction Earthquake Cycles, by Kelin Wang 18. Distinct Updip Limits to Geodetic Locking and Microseismicity at the Northern Costa Rica Seismogenic Zone: Evidence for Two Mechanical Transitions, by Susan Y. Schwartz and Heather R. DeShon Part VI. Regional Scale Deformation 19. Collision Versus Subduction: From a Viewpoint of Slab Dehydration, by Tetsuzo Seno 20. Subduction and Mountain Building in the Central Andes, by Jonas Kley and Tim Vietor List of Contributors Index

Abstract Sands penetrated at shallow depths (< 1 km) in the Gulf of Mexico can flow to the surface and undermine seafloor installations. Here we explore aspects of natural fluid expulsion structures from these "shallow-water sands". Highresolution seismic data shows no obvious sand injection structures from the Blue Unit, a shallow-water sand, in the Ursa area, but perhaps nascent dike formation. Also, channel sands stratigraphically above the Blue Unit are overpressured, with a possible pressure source in the latter. In the Ursa area the maximum principal stress is vertical and constrained by the integrated unit weight of the sediments at IODP Site U1324. The stress necessary to form hydrofractures (the minimum principal stress plus negligible tensional strength of the sediment) is about 85% of the maximum principal stress indicating an approximately isotropic state of stress. High Poisson's ratios (0.46-0.49), determined from wireline compressional and shear wave velocities at Site U1324, are also consistent with a nearly isotropic state of stress at shallow depths. Estimated maximum fluid pressures are about 80% of the total overburden stress, less than necessary to hydrofracture the sediment. Therefore, the fluid pressure is insufficient to form sand injectites. The formation of sand injectites, as observed in the geologic record, may require very high fluid pressures, approaching lithostatic values. Diverse orientations of ancient injectites support the necessity of very high fluid pressures and a nearly isotropic state of stress. Introduction Drilling operations in the Gulf of Mexico, especially in deep water, have encountered unconsolidated and overpressured sand at shallow depths. Once penetrated these sands flow into the wells, washout, erode around the exterior of casing strings, and locally flow to the surface [1]. Surface features associated with this ?shallow water flow' behavior include cracks, faults, and gully-like features [2]. In areas where this behavior has occurred, the seafloor appears to be undergoing subsidence associated with sediment loss at depth. In the Ursa drilling area (Figure 1) the shallow-water sands are overpressured which initiates the flow into the boreholes [1, 3]. The shallow-water sands in the Ursa area readily flow into boreholes and to the surface if the borehole is normally pressured (i.e. the hole is uncased or an underbalanced mud is used). Therefore, an obvious question is whether there are any natural injections from the shallow-water sands or other evidence of fluid flow. Given the overpressured nature of the shallow-water sands [3], naturally occurring injections and other fluid migration features could be expected. These "injectites" are common in some petroleum systems where they complicate sand reservoir geometry and have important implications for hydrocarbon production [4]. Because fluid pressure and sand geometries are well-constrained, the Ursa area may be an important natural laboratory to ascertain the limiting fluid pressure and stress conditions for the occurrence of injectites and other fluid flow features. Fluid Migration and Sand Injection from Blue Unit Sand in IODP Drilling Area During the last 70 ka, rapid deposition, directly south of the Mississippi River accumulated a sand and mud sequence, the Blue Unit [5].

The Yellow Bank creek complex (YBCC) is a large, upper Miocene injectite complex, one of numerous injectites northwest of Santa Cruz, California. The feeder for these injectites is the Santa Margarita Sandstone, a shelfal sandstone unit that is also the reservoir rock in several exhumed oil fields. The impermeable cap rock for these oil fields, the Santa Cruz Mudstone, was breached by sand injectites, some of which reached the sea floor. Located near the edge of one of these oil fields, the YBCC is a dike-sill complex that shows evidence for multiple phases of injection by fluidized sand that was initially gas or water saturated and later possibly oil bearing. Vertical injection of a large sand dike along a fracture was followed by lateral injection of a sill from the dike along bedding planes in the Santa Cruz Mudstone. Flow differentiation during injection of fluidized sand into the sill formed centimeter-scale layering in its lower part. Subsequent emplacement of oil into this sand may have occurred by injection and by seepage that displaced pore water, producing sand masses that became preferentially cemented by dolomite. Some evidence suggests that the injection and cementation occurred at relatively shallow burial depths beneath the sea floor, with the injection resulting from a combination of possible seismic shaking and migration of overpressured fluids from more deeply buried parts of the Santa Margarita Sandstone. A pervasive lamination marked by limonite staining developed following uplift and subaerial exposure of the complex, possibly in a groundwater environment.

We have investigated the fabric and the deformational processes of an exhumed subduction zone thrust active at seismogenic depths. The Rodeo Cove thrust zone, which outcrops north of the Golden Gate Bridge of San Francisco, imbricates two basalt-chertsandstone sequences belonging to the Marin Headlands terrane (Franciscan Complex). The thrust outcrop is a 200-m-thick complex zone that displays a range of stratal disruption from incipient deformation to a broken formation in the central part of the outcrop, dominated by basaltic lithologies, where zones of concentration of deformation have been mapped. Disruption is made by variably dense discrete fault systems synthetic to the main thrust (R and P fractures). These faults are marked by cataclasites with a shaly matrix that shows a scaly foliation defi ned by chlorite and pumpellyite, which also constrain the depth of faulting (8‐10 km, T = 200‐250 °C) within the seismogenic zone. The central part of the fault also features the densest system of carbonate-fi lled veins. Veins occur in the broken formation matrix and fragments, in both cases parallel to the foliation. The veins are either folded, truncated, or pressure-solved along the cleavage. Cementation and hardening of shear surfaces of the fault core may have caused the distribution, as opposed to localization, of subsequent slip events. The fault core may have developed in basaltic rocks because of their inherently high permeability and propensity to transmit overpressure from deeper levels of the subduction zone. Our analysis has shown that accretionary deformation is strongly controlled by injection of overpressured fl uids occurring through systems of multiple dilatant fractures grossly parallel to the decollement zone. The crosscutting relationships between veining and foliation suggest that fl uid injection is cyclic and, consequently, that large transient variations in permeability and cohesion may occur. The repeated injection of veins parallel to the fault zone may be explained by cyclic changes of the stress, or by difference in tensional strength parallel to and perpendicular to the foliation, both of which would require extremely high fl uid pressure. We interpret the features of the Rodeo Cove thrust zone as evidence of the seismic cycle and hypothesize a compressional stress fi eld in the interseismic phase and an extensional stress fi eld in the immediately postseismic phase.

In the Kodiak accretionary complex, Kodiak Island, Alaska, pseudotachylyte occurs in black, locally vitreous ultrafine-grained fault rock. Microscopic observations show that the pseudotachylytes are composed of glass, with vesicles, amygdules, microlites, and flow structures, indicating a frictional melt. The pseudotachylyte is gradational to cataclasite and shows outcrop-scale injection and ductile deformation structures. The cataclasite was ductily mobile (i.e., fluidized) simultaneous with the formation and emplacement of pseudotachylyte melt. The pseudotachylytic rocks postdate the stratal disruption fabric of associated shear-zone melanges and show similar direction of thrust transport, and have undergone limited subsequent deformation. We interpret the stratal disruption as resulting from underthrusting of the subducting plate and pseudotachylyte development as the final activity of this thrust surface. The gradational contacts between pseudotachylyte and cataclasite demonstrate that the cataclasite also formed as a seismic product and may represent paleoseismic rupture zones, possibly of very great earthquakes, with or without accompanying pseudotachylytes. The pseudotachylytes are voluminous, and many are spatially disconnected from generation surfaces. This style is distinct from pseudotachylytes described in other environments, and this may explain the rarity of documented examples of subduction-thrust pseudotachylyte.

(1) Along representative cross sections of the Apennines and the Northern Barbados accretionary prisms, we measured the area, the decollement depth, the angle a of the upper envelope and the angle b of the dip of the regional monocline. The continental sections of the Apennines accretionary prism have a deeper decollement than the oceanic sections of the Northern Barbados, 6-10 km depth and

An exceptional exposure of the San Gregorio Fault provides the opportunity for detailed observations of structural fabrics within an active fault zone. Where it is exposed in the intertidal zone in Moss Beach, California, the San Gregorio Fault juxtaposes different sedimentary lithologies within the Pliocene Purisima Fm. An approximately 10 meter-wide zone of clay-rich foliated gouge marks the fault. The damage zone is approximately 100 meters wide, and the distribution of deformation is heterogeneous across the fault zone. Structural fabrics in the northeast fault block include breccias and both microscopic and outcrop-scale shear zones; these record the effects of cataclasis on porous sandstones and conglomerates. Deformation in the mudstones of the southwest fault block is accommodated by an incipient scaly foliation as well as by numerous fractures and faults. Microstructural analyses indicate that the San Gregorio accommodates dextral strike-slip offset as well as a component of west-side up reverse motion. Evidence for the role of fluids in this fault zone includes field relations and geochemical data. Anomolous hydrocarbon content within the foliated fault gouge indicate that the fault is a migration conduit. Fluctuations in fluid pressure within this fault zone may help elucidate the mechanics and seismogenic potential of the San Gregorio Fault.

Sandstone intrusions are widespread west of Santa Cruz, California and were emplaced during late Cenozoic tectonic deformation of this region. Among these is a very large and complex intrusion which is well exposed along the coastline at Yellow Bank Creek. Here, fluidized sands from the Miocene Santa Margarita Sandstone were injected upward into fractured biosiliceous rocks of the Santa Cruz Mudstone, probably due to faulting and seismic shaking. The complicated internal structure of this intrusion includes sedimentary xenoliths, fluidization structures, and secondary limonite staining. The latter likely occurred during oxidation by groundwater and produced conspicuous, complicated layering which serves to mask and confuse interpretations of the earlier-formed features. Among the earlier formed features are fluidization structures, comprising (1) flow banding which records injection of sands horizontally in silllike areas of the intrusion, and (2) heave structures which reflect mainly vertical injection of hydrocarbons and sands partially saturated with hydrocarbons into water-saturated sands. This latter type of injection appears to have occurred at a hydrocarbon front that was derived from either a localized petroleum accumulation or else from remnants of hydrocarbons that had mostly migrated updip prior to the clastic intrusion event. Dolomitic cementation occurred preferentially in the hydrocarbon-saturated sands due to degradation of the hydrocarbons. Paleotemperature estimates of the intrusive sandstones (by apatite fission track analysis) and of the host Santa Cruz Mudstone (by vitrinite reflectance) indicate maximum temperatures of about 60°C for the former, 50°C for the latter. Our data suggests that initial fluidization began in water-saturated sands of the bioturbated facies in the Santa Margarita Sandstone; following upward intrusion of these sands, fluidization and injection expanded into hydrocarbon-bearing sands within the cross-bedded facies of the same unit.