Pseudotachylyte Layers Research Articles

Postseismic fluid discharge chemically recorded in altered pseudotachylyte discovered from an ancient megasplay fault: an example from the Nobeoka Thrust in the Shimanto accretionary complex, SW Japan

Megasplay fault branching from plate boundaries of subduction zones is thought to be important sources of earthquakes and tsunamis. In this study, we performed structural and geochemical analyses on a fossilized megasplay fault (the Nobeoka Thrust of the Shimanto accretionary complex) to understand fluid-rock interaction and how the splay fault plays a role in fluid flow in the seismogenic zone. As a result of structural observations, we report that the principal slip zone (PSZ) of the Nobeoka Thrust is composed of foliated cataclasite originating from a sandstone-shale mélange and includes a thin (~ 1.5 mm thick) pseudotachylyte layer. Major and trace element composition analysis and EPMA element mapping revealed that the pseudotachylyte is enriched in Li and Cs within the PSZ, as well as in the slip zone of a minor fault in the footwall. Li and Cs enrichment in pseudotachylyte is interpreted as a result of fluid-rock interaction in the postseismic stage, because such an anomaly only results from a large fluid/rock ratio (R > 512–24 at 250–350 °C) under the influence of Li- and Cs-enriched fluids. The amount of fluid that reacted with the pseudotachylyte is estimated to be 1.78 × 100 to 7.61 × 103 m3.

Distributed deformation along the subduction plate interface: The role of tectonic mélanges

Recent geophysical monitoring of subduction zones has unraveled a complete spectrum of plate coupling behaviors, from coupled portions rupturing during earthquakes to decoupled portions slipping aseismically. However, the deformation mechanisms and the exhumed rock corresponding to these contrasted behaviors are not yet identified. Tectonic mélange zones are thought to play a major role in the deformation of the plate interface as they represent remnants of the subducted plate scraped off by the overriding plate. In this work we examine several tectonic mélange zones (Hyuga, Okitsu, Mugi) from the Shimanto Belt, an accretionary prism in southwest Japan connecting to the active Nankai subduction zone. These tectonic mélange zones have a block-in-matrix structure, with lenses of sandstones and basalts within a metapelitic matrix, and their deformation is distributed over zones of hundreds of meters in thickness. In addition, the examples of mélange considered here are bounded by sharp faults, some of them bearing pseudotachylyte layers, so that distributed deformation within the mélange and localized deformation on its boundary are juxtaposed. Distributed deformation involves the development of a foliation, as well as of a pervasive network of macroscopic and microscopic shear zones. Along with slip on this network, strain proceeds by fracturing and precipitation of quartz, in long veins parallel to the foliation or in smaller cracks perpendicular to stretching and forming in the neck of competent lenses of sandstones or former quartz veins. The analysis of shear band kinematics shows in all three examples a dominant, top-to-the-trench sense of shear, consistent with deformation along the plate boundary during subduction. Moreover, most shear zones, when foliation is restored back to syn-subduction position, are extensional structures. Finally, the geometry and kinematics of the mélange-bounding faults, as well as radiometric constraints, show that in most cases the faults (=localized structures) were formed during a later stage than mélange internal deformation. These findings bear several consequences on the structure and dynamics of the subduction plate boundary at seismogenic depths. First, there is no support for a model of plate boundary fault zone composed simultaneously of localized slip zones and domains of more distributed deformation. Second, rather than proposed models of underplating, where all deformation is localized into the thrusts bounding the tectonic sheets, we suggest that underplating was to a large extent accommodated by distributed deformation within the mélange sheets. This underplating model accounts for (i) the large amount of strain within the mélange, (ii) the absence of contractional structures during underplating, (iii) thinning of the mélange required by the network of extensional shear bands and stretched boudins. Third, mélanges appear as likely candidates for portions of the plate interface deforming by aseismic slip. The seismic vs. aseismic character of the plate interface might depend on the ability of sediments on top of the subducting plate to undergo distributed strain, which in turn depends on the efficiency of pressure solution to operate.

Slip parameters on major thrusts at a convergent plate boundary: regional heterogeneity of potential slip distance at the shallow portion of the subducting plate

Understanding variations of slip distance along major thrust systems at convergent margins is an important issue for evaluation of near-trench slip and the potential generation of large tsunamis. We derived quantitative estimates of slip along ancient subduction fault systems by using the maturity of carbonaceous material (CM) of discrete slip zones as a proxy for temperature. We first obtained the Raman spectra of CM in ultracataclasite and pseudotachylyte layers in discrete slip zones at depths below the seafloor of 1–4 km and 2.5–5.5 km, respectively. By comparing the area-under-the-peak ratios of graphitic and disordered bands in those Raman spectra with spectra of experimentally heated CM from surrounding rocks, we determined that the ultracataclasite and pseudotachylyte layers had been heated to temperatures of up to 700 and 1300 °C, respectively. Numerical simulation of the thermal history of CM extracted from rocks near the two slip zones, taking into consideration these temperature constraints, indicated that slip distances in the ultracataclasite and pseudotachylyte layers were more than 3 and 7 m, respectively. Thus, potential distance of coseismic slip along the subduction-zone fault system could have regional variations even at shallow depth (≤ 5.5 km). The slip distances we determined probably represent minimum slips for subduction-zone thrusts and thus provide an important contribution to earthquake preparedness plans in coastal areas facing the Nankai and Sagami Troughs.

Identifying deformed pseudotachylyte and its influence on the strength and evolution of a crustal shear zone at the base of the seismogenic zone

On a strand of the Norumbega fault system, a Paleozoic, subvertical, seismogenic fault system in northeastern New England, USA, we document changes associated with the formation and deformation of pseudotachylyte to form ultramylonite/phyllonite layers. We consider how those textural and mineralogical changes affected the rheology of the layer and how significant volumes of pseudotachylyte over time may have weakened the shear zone.The Norumbega fault system is characterized by a number of mylonitic shear zones exhumed from depths of ~10–15km, some of which preserve evidence for mutually-overprinting pseudotachylyte and mylonite. Along one of these, the Sandhill Corner shear zone, all stages of the pseudotachylyte to ultramylonite/phyllonite transformation are preserved, from (1) primary pseudotachylyte structures to (2) initial mineral crystallization, (3) grain coarsening and reactions, and (4) viscous deformation. Our observations show that ultramylonite layers exhibit identifying features that when present together are distinctive of a pseudotachylyte origin. Using these features, we estimate that ~5–50% of the rock volume in the Sandhill Corner shear zone (mean ~30%, locally >50%) is deformed pseudotachylyte, suggesting that deformed pseudotachylyte may be more prevalent than previously thought in faults exhumed from the base of the seismogenic zone.The Sandhill Corner shear zone localized along the contact between two rheologically-contrasting units. Mylonite fabric intensity and the occurrence of fresh and deformed pseudotachylyte increase with proximity to the contact and shear zone core, indicating that seismic rupture also localized there. A decrease in grain size promoting grain-size-sensitive creep and a progressively interconnected mica network associated with local basal slip within the deformed pseudotachylyte both worked to decrease the strength of those layers. The formation of multiple generations of weak, deformed pseudotachylyte layers at or near the lithologic contact may have played an important role in the spatiotemporal persistence of the shear zone core.

Analyses of pseudotachylyte from Hole-B of Taiwan Chelungpu Fault Drilling Project (TCDP); their implications for seismic slip behaviors during the 1999 Chi-Chi earthquake

The seismic slip behavior during the 1999 Chi-Chi, Taiwan, earthquake ( M w 7.6) was contrastive between the northern and southern segments of the activated Chelungpu fault; large, fast and smooth slips with large stress drop in the north, while smaller, slower and irregular slips with smaller stress drop in the south. We analyzed the pseudotachylyte samples recovered from 1194 m, 1243 m and 1314 m depths of Hole-B of Taiwan Chelungpu fault Drilling Project (TCDP) to reveal the spatial difference in friction mechanism. All pseudotachylyte layers are thin (0.7–2.8 cm), the volume fraction of protoliths is very large (more than 63%), and the estimated temperature distribution is very heterogeneous from ca. 750–1750 °C. These observations suggest that these pseudotachylyte melts were in the partial melting regime of Montgomery [Montgomery, R.S., 1976. Friction and wear at high sliding speeds. Wear 36, 275–298] where friction coefficient is abnormally large. Similar pseudotachylyte was found already in the core sample from 175 m depth of the Nanto borehole penetrating the southern fault. Since both pseudotachylyte samples from the two boreholes are older than the 1999 Chi-Chi event and have been uplifted from depths farther down-dip of their current locations, it is likely that recent seismic ruptures also would have encountered these mechanical barriers of viscous melt patches at deeper parts in the north than in the south. Elastohydrodynamic lubrication of clayey gouge worked effectively at the shallower parts of the northern segment, however there is no evidence that it played an important role in the south. These differences are the plausible causes of the contrastive local slip behaviors during the Chi-Chi earthquake.

Probing fault zone heterogeneity on the Nojima fault: Constraints from zircon fission-track analysis of borehole samples

Fission-track (FT) thermochronologic analysis was performed on zircon separates from rocks in and around the Nojima fault, which was activated during the 1995 Kobe earthquake. Samples were collected from the University Group 500 m (UG-500) borehole and nearby outcrops. FT lengths in zircons from localities > 25 m away from the fault plane as well as one 0.1 m away from the fault in the footwall are characterized by concordant mean values of ∼ 10–11 μm and unimodal distributions with negative skewness, which showed no signs of appreciable reduction in FT length. In contrast, those adjacent (< 3 m) to the fault at depths on the hanging wall side showed significantly reduced mean track lengths of ∼ 6–8 μm and distributions having a peak around 6–7 μm with rather positive skewness. The former pattern is interpreted to reflect cooling through the zircon partial annealing zone (ZPAZ), without later, partial thermal overprints. The latter indicates substantial track shortening due probably to secondary heating by a thermal event(s) that locally perturbed the geothermal structure. Modeled zircon FT length and age data of partially annealed samples from the UG-500 borehole revealed a cooling episode in the ZPAZ that started at ∼ 4 Ma within ∼ 3 m from the fault plane, whereas those from the Geological Survey of Japan 750 m borehole record cooling started at ∼ 31–38 Ma within ∼ 25 m from the fault. On the basis of one-dimensional heat conduction modeling as well as the consistency between the degree of FT annealing and the degree of deformation/alteration of borehole rocks, these cooling ages in both boreholes are interpreted as consequences of ancient thermal overprints by heat transfer or dispersion via fluids in the fault zone. Together with the zircon FT data of a pseudotachylyte layer recently analyzed, it is suggested that the present Nojima fault system was reactivated in the Middle Quaternary from an ancient fault initiated at ∼ 56 Ma at mid-crustal depths. Also shown is a temporal/spatial variation in terms of the thermal anomalies recorded in the fault rocks, implying heterogeneity of hot fluid flows in the fault zone.

Numerical simulations of frictional melting: Small dependence of shear stress drop on viscosity parameters

Frictional melting during coseismic slipping strongly affects dynamic sliding. An idealized numerical simulation of shear stress drop due to lubrication of fault surfaces by frictional melting is performed to investigate dynamical adjustment of melt thickness and its dependence on material properties. We numerically solved one‐dimensional heat conduction equations coupled with hydrodynamic equations of motion of the melt layer. We adopt a simple Arrhenius‐type temperature dependence of melt viscosity with a parametric factor η0. After the initial transient stage, the evolution of the melt layer asymptotes to the late stage, where the melt thickness increases as and the generated shear stress drops as 1/. An analytic self‐similar solution of the temperature profile for the late stage is obtained. The balance between the viscous heat generation and conductive heat loss characterizes the late stage. For small η0, the melt thickness is proportional to η0, whereas it will saturate for a larger η0. This is because the strong temperature dependence of viscosity enables an automatic adjustment of viscosity, which drops through the temperature increase due to the viscous heating even if a large η0 is chosen. Numerical results and the thickness data of natural pseudotachylyte layers were compared. A maximum ratio X/ (X is thickness, D is sliding distance) exists above which no solution is found. The effective thermal conductivity of the melt layer should be large, probably due to the sliding surface roughness. Comparison with laboratory experimental results showed that the normal stress applied to the sample is an important parameter for stress drop.

Occurrence of mylonite zones and pseudotachylyte veins around the base of the upper crust: An example from the southern Hidaka metamorphic belt, Samani area, Hokkaido, Japan

Geological studies on exhumed pseudotachylyte-bearing mylonite zones in S-type tonalite were carried out in the southern Hidaka metamorphic belt, Hokkaido, Japan. Mylonitization is characterized by (1) development of composite planar fabrics, (2) grain size reduction, and (3) change in modal composition an increase in mica content and a decrease in quartz content from protolith to mylonite. Mylonite zones are heterogeneously concentrated in the host rocks. At microscopic scales, shear deformation is concentrated heterogeneously in finegrained layers along C-surfaces. Most of the pseudotachylyte layers are subparallel to the C-surface, and tend to overprint thick mylonite zones. The heterogeneous development of mylonite zones, which may be activated as layers of co-seismic slip, should be incorporated into numerical modeling of seismogenic zones.

Dating pseudotachylyte of the Nojima fault using the zircon fission‐track method

The zircon fission‐track analysis of pseudotachylyte answered how old the Nojima fault is, which was activated during the 1995 Kobe earthquake in Japan. The time of seismic fault motions at the crustal interior can be constrained by dating pseudotachylyte formation. We succeeded in separating zircons from the pseudotachylyte layer of the fault as well as in dating them by the fission‐track method. Here we show that the layer has a significantly younger age (56 ± 4 Ma) than the initial cooling age of the host rock, suggesting that the fault was already initiated at ∼56 Ma, much older than the previous estimate based on the spatial distribution of sediments around the fault.

Fluidization and melting of fault gouge during seismic slip: Identification in the Nojima fault zone and implications for focal earthquake mechanisms

The record of physical processes that occur during seismic slip events is well preserved in fault rocks from the active Nojima fault in Japan. The fault rocks formed at about 3 km depth, and comprise thin alternating layers of very fine gouge and pseudotachylyte derived from granite. Each layer is thinner than a few millimeters, and corresponds to one seismic slip event. The very thin slip zone width suggests that some mechanisms of slip weakening operated, and our studies of the fault rocks suggest that fluidization and melting of gouge were particularly important. Fluidized and nonfluidized gouges were distinguished using the detection probability of fragmented counterparts method. It is known from granular material science that the phase transition from a grain friction regime to fluidization of granular materials can occur only by a very small decrease in volume fraction of solid grains. Once fluidization occurs, the frictional resistance decreases abruptly to nearly zero even before thermal pressurization reaches its extreme state. When fault gouge is melted, frictional resistance is governed by the viscosity of melt which depends mainly on temperature, chemical composition, H2O concentration and the volume fraction of unmelted grains. For each pseudotachylyte layer, we estimated the temperature of melt using various temperature indices, and measured volume fraction of unmelted grains. We synthesized these data to reconstruct the change in viscosity during seismic slip events. The melt viscosity is very high (107–9 Pas) during initial melting due to the combined effect of low temperature (750°–800°C) and large volume fraction of solid grains. Thus seismic slip is inevitably restrained instantaneously. Once this mechanical barrier is overcome, the viscosity reduces continuously and dramatically as slip increases. At 1000°C, the viscosity reduces to 104Pas and eventually to 10Pas at 1280°C. Thus stress drops almost completely and rupturing tends to run away.

Fluid inclusions in pseudotachylytes from the Nojima fault, Japan

Pseudotachylytes (i.e., rocks formed by frictional melting) have been observed in the Nojima fault that was penetrated by the Hirabayashi borehole drilled 1 year after the 1995 Hyogo‐ken Nanbu (Kobe) earthquake. These rocks display millimeter‐scale banding defined by different pseudotachylyte layers. The nature of unmolten crystal fragments (K‐feldspar, albite, calcite, and/or quartz) allows us to infer a minimum melting temperature of 1200°C. The glass has a 8±3% volatile content and a higher CaO content than that of the parent granodiorite, thus suggesting that pseudotachylyte formation occurred in an already altered, calcite‐bearing and hydrated fault zone. Fluid inclusions have been observed in the glass and are filled with a dense low‐salinity (4.3±1.2 wt % eq. NaCl) CO2‐H2O fluid characterized by steep isochoric curves. The intersection between the isochores and the measured 24°C/km geothermal gradient indicates a minimum 15 km depth for the pseudotachylyte formation. Such a depth suggests that an important uplift has brought the studied pseudotachylytes to their present position and that these pseudotachylytes are probably pre‐Miocene in age. The calculated thermal evolution of a typical millimeter‐scale pseudotachylyte layer indicates that cooling does not last more than a few seconds. This indicates that related seismic processes, such as deceleration of fault movement and healing of the fault, were both very rapid. The geometry and thermal budget of the millimeter‐thick pseudotachylytes suggest that seismic events of magnitude 6 to 7 are responsible for their formation.

Size distribution of clasts in experimentally produced pseudotachylytes

High-velocity frictional experiments on monzodiorite, at sliding rates of 1.3 m/s and at normal stresses of 2 MPa, produced dark 100 μm thick pseudotachylyte layers on the sliding surfaces. The experimentally produced pseudotachylyte contains clasts of various sizes in an ultrafine matrix, similar to the texture observed for natural pseudotachylytes. The size distribution of clasts in the experimental pseudotachylyte obeys a power law of the form: N= N′(1+ r/ r′) − D , where N is the cumulative number of clasts with a size greater than r, r′ and D are material constants, and N′ is a constant that depends on the total number of measurements. The total area of ultrafine-grained clasts in a thin section, estimated from the size distribution is less than 0.5%, whereas the actual area of ultrafine matrix is measured as 60–70% of the total product. This suggests that the matrix cannot be regarded as a product of only comminution processes, consistent with natural pseudotachylytes, and that the ratio of actual area of the ultrafine-grained matrix to the hole product represents almost the ratio of the melt (glass). This estimate is consistent with the ratio of glass determined from an X-ray diffraction analysis.

Generation of Pseudotachylyte by Ancient Seismic Faulting

Summary Pseudotachylyte occurs as vein material infilling highly brittle shear and extensional fractures developed along the western margin of the late Caledonian, Outer Hebrides Thrust zone in NW Scotland. Vein geometries and textures show clearly that the pseudotachylyte has been through a melt phase. From the composition of the pseudotachylyte matrix which is close to that of a basaltic andesite, probable melt temperatures of around 1100°C are inferred. Field and theoretical studies demonstrate that the pseudotachylyte was generated by relatively high stress seismic faulting in crystalline sialic crust devoid of an intergranular fluid, most probably at the time of thrust inception and at a depth of around 4–5 km. A study of pseudotachylyte-bearing ‘single-jerk’ microfaults shows that the slip (d) is related to the thickness of the pseudotachylyte layer (a) by the equation, where d and a are measured in centimetres. Work-energy calculations based on this empirical relationship suggest that the pre-failure shear stress on the microfaults must have been as high as 1.6 kbar to overcome the initial frictional resistance (τf), which decreases with increasing slip during a single movement according to the relationship, which may arise solely from the viscous shear resistance of the melt layer. Delineation of palaeoseismic zones by the recognition of those cataclastic rocks which are necessarily the products of earthquake faulting, may assist in the determination of ancient plate boundaries.