Nojima Fault Zone Research Articles

Absolute Dating of Fault-Gouge Material Using Isothermal Thermoluminescence: An Example from the Nojima Fault Zone, SW Japan

Establishing the absolute age of palaeoearthquakes is of great significance for the assessment of the seismicity and seismic hazards of a region. As such, several different geochronological techniques to date earthquake-related material have been developed to provide answers on the time of past earthquakes. The present study is part of a wider palaeoseismic research project conducted in the Nojima Fault Zone (NFZ), where the 1995 Mw 6.9 Kobe (Japan) earthquake was triggered, to assess the suitability of the isothermal thermoluminescence (ITL) dating technique on fine-grained quartz and medium-grained feldspar and to provide a sequence of ages for fault-rock samples separated from a drilled core that was retrieved from a depth of ~506 m. Our analysis reveals that ITL can produce consistent dating results and can be considered a reliable luminescence technique for the absolute dating of fault-gouge material. The produced ITL ages signified the existence of repeated seismic events within the NFZ that took place through the late Pleistocene period, with gouge ages spanning from 78.6 ± 4.2 to 13.4 ± 1.4 ka; however, overestimation of the produced ITL dating results may be apparent. Nonetheless, even though some degree of overestimation is considered, ITL dating results denote the oldest possible age boundary of formation (or luminescence signal resetting) of the collected fault-gouge layers.

Absolute Dating of Past Seismic Events Using the OSL Technique on Fault Gouge Material—A Case Study of the Nojima Fault Zone, SW Japan

AbstractThis study is the first part of a comprehensive paleoseismic study in the Nojima Fault Zone (NFZ), the seismogenic fault of the 1995 Mw 6.9 Kobe (Japan) earthquake, which explores the use of the luminescence dating method for understanding past faulting time and assessing the activity of faults developed in basement rocks. Our approach is focused on methodological aspects of the optically stimulated luminescence (OSL) technique and reports on a series of ages of fault gouge samples. Our analysis revealed that (1) quartz OSL signals have been reset (at least partially) during past seismic faulting events, (2) fault gouge layers bounded by sharp fault planes are younger than gouge layers far from the main fault planes, and (3) the most recent seismic faulting, affecting the different gouge layers on the Nojima fault took place from 62.8 ± 4.3 ka to 18.5 ± 1.3 ka, while on the Asano fault from 139.3 ± 8.9 ka to 45.8 ± 3 ka, with these ages representing maximum possible ages. In this regard, this is the first successful absolute dating attempt on the NFZ using quartz grains and the only luminescence study, as yet known, producing a series of ages representing neotectonic activity of the Nojima and Asano faults during the late Pleistocene and middle to late Pleistocene, respectively. The present work has indicated that OSL is potentially a promising technique for dating fault gouges and assessing the activity of faults, although more work is needed for further refinement.

Fractures and Subsidiary Faults Developed in the Active Strike‐Slip Nojima Fault Zone, Japan, and Tectonic Implications

AbstractThe Nojima Fault Zone, on which the 1995 Mw 6.9 Kobe (Japan) earthquake occurred, is characterized by pulverized cataclastic rocks with numerous fractures and subsidiary faults. The fractures and subsidiary faults were observed in trench walls and using borehole televiewer (BHTV) images from boreholes NFD‐1 (~1,000‐m depth) and AFD‐1 (~700‐m depth), which were drilled through the Nojima Fault and its branch, the Asano Fault, respectively. Measurements show that the orientations of fractures and subsidiary faults in the two boreholes are concentrated at N10–70°E, averaging N40°E. In contrast, the orientations of fractures and subsidiary faults in the walls of a trench across the main surface trace of the Asano Fault are constrained to N0–40°E, averaging N20°E, parallel to the general trend of the Asano Fault but different with that observed in the deep holes at the Ogura site. Structural analysis shows that the fractures and subsidiary faults are mainly concentrated in Riedel Y and R shears of the right‐lateral strike‐slip Nojima Fault and its branch Asano Fault. Our findings demonstrate that the fractures and subsidiary faults within the Nojima Fault Zone were formed by dextral strike‐slip shearing under the current ENE‐WSW to E‐W compressive regional tectonic stress field, related to ongoing subduction of the Pacific and Philippine Sea plates beneath the Japanese Islands.

TEM microstructural analysis in a fault gouge sample of the Nojima Fault Zone, Japan

A fault gouge sample, recovered from 389m depth of the Disaster Prevention Research Institute (DPRI) 500m bore hole of the Nojima fault, which triggered the 1995 M7.3 Kobe earthquake, was examined with transmission electron microscopy (TEM). We observed different types of amorphous material containing tiny flakes of crystalline phases. TEM-EDX analysis of the matrix indicates a pure smectite composition suggesting that matrix and crystalline flakes have the same composition. The detection of tiny crystalline flakes in the amorphous material may confirm previous assumption from the alteration of pseudotachylyte into smectite in fault rocks.

Reduction in BET surface area of Nojima fault gouge with seismic slip and its implication for the fracture energy of earthquakes

A common view concerning the energetics of seismogenic fault motion is that at least part of the fracture energy is consumed in grain crushing in the fault zone, and that this part may be estimated by grain-size analysis of fault rocks. We address this problem by conducting room-dry friction experiments on Nojima fault gouge at subseismic to seismic slip rates (0.009–1.31m/s) and at normal stresses up to 3.64MPa, and by measuring the BET surface area of the gouge before and after the experiments. Where it cuts granite, the Nojima fault zone has BET surface area of about 65×106m2 per unit fault area (1m2). Clayey and granular fault gouges, composed mainly of quartz, plagioclase, kaolinite and smectite, were collected from a granitic fault zone at a new outcrop in Funaki, Awaji Island, southwest Japan. Both clayey and granular gouges exhibit dramatic weakening at high slip rates. Specific BET surface areas of clayey and granular gouges decreased with increasing slip rate from 46.0 and 15.4m2/g before deformation to about 20 and 5m2/g after deformation (55–70% reduction), respectively. Microstructural observations revealed that grain welding within the slip zones at high slip rates reduced grain surface area. The energetics of seismic fault motion should be examined with broader views taking into account grain crushing, grain welding, decomposition and frictional melting.

Timing of brittle deformation within the Nojima fault zone, Japan

K–Ar ages from clay fault gouges of outcrop and core samples (UG500 well) from the Nojima fault zone, Awaji Island, Japan are consistent internally with established field constraints, and with apatite and zircon fission track ages, demonstrating the suitability of this method for dating of brittle deformation. Six unspiked and thirty conventional K–Ar data of clay minerals separated from gouge zones were analysed. Characterized by protoliths of high-grade metamorphic and magmatic rocks, the samples from the Nojima fault zone offer the opportunity to distinct identify newly grown illite in the fault planes. All dated sample fractions were characterized by XRD, SEM and TEM. K–Ar ages of 29 separates have an age range from 63.4±1.3Ma (Early Palaeocene–Danian) to 42.2±1.0 (Palaeogene–Middle Eocene–Lutetian). Some <0.1 and <0.4μm fractions of samples in close proximity to a pseudotachylyte zone are thermally influenced and indicate loss of radiogenic Ar. The K–Ar data from illite support a model that the Nojima fault zone was initiated ∼55Ma ago based on ZFTA data. The data confirm elevated temperatures and a heterogeneous thermal history within the study area. The <2μm ages of the outcropping samples document the age of gouge formation, whereas the <0.1 and <0.4μm illite fractions suggest the influence of a secondary thermal heating event probably caused by circulation of hot fluids within the fault zone about 31–38Ma ago and even a potential influence of Quaternary faulting. Ar diffusion models support elevated temperatures for outcrop samples collected near the Hirabayashi site. The thermal anomalies documented in the outcrop samples could not be observed in the UG500 core samples supporting further a heterogeneous thermal history model within the Nojima fault zone.

Geometry of the Nojima Fault at Nojima-Hirabayashi, Japan – II. Microstructures and their Implications for Permeability and Strength

Samples of damage-zone granodiorite and fault core from two drillholes into the active, strikeslip Nojima fault zone display microstructures and alteration features that explain their measured present-day strengths and permeabilities and provide insight on the evolution of these properties in the fault zone. The least deformed damage-zone rocks contain two sets of nearly perpendicular (60–90° angles), roughly vertical fractures that are concentrated in quartz-rich areas, with one set typically dominating over the other. With increasing intensity of deformation, which corresponds generally to increasing proximity to the core, zones of heavily fragmented rock, termed microbreccia zones, develop between prominent fractures of both sets. Granodiorite adjoining intersecting microbreccia zones in the active fault strands has been repeatedly fractured and locally brecciated, accompanied by the generation of millimeter-scale voids that are partly filled with secondary minerals. Minor shear bands overprint some of the heavily deformed areas, and small-scale shear zones form from the pairing of closely spaced shear bands. Strength and permeability measurements were made on core collected from the fault within a year after a major (Kobe) earthquake. Measured strengths of the samples decrease regularly with increasing fracturing and fragmentation, such that the gouge of the fault core and completely brecciated samples from the damage zone are the weakest. Permeability increases with increasing disruption, generally reaching a peak in heavily fractured but still more or less cohesive rock at the scale of the laboratory samples. Complete loss of cohesion, as in the gouge or the interiors of large microbreccia zones, is accompanied by a reduction of permeability by 1–2 orders of magnitude below the peak values. The core samples show abundant evidence of hydrothermal alteration and mineral precipitation. Permeability is thus expected to decrease and strength to increase somewhat in active fault strands between earthquakes, as mineral deposits progressively seal fractures and fill pore spaces.

Healing microstructures of experimental and natural fault gouge

The healing of fault gouge was studied by examining microstructures of naturally and experimentally produced granitoid fault rock. We performed deformation experiments on intact granitoid rock samples at T = 300–500°C, Pc = 500 MPa, and = 1.2 × 10−4 – 1.3 × 10−7 s−1 with 0.2 wt% H2O added. Healing experiments were carried out on deformed samples at T = 200–500°C, Pc = 500 MPa, for 4 h to 14 days under hydrostatic and nonhydrostatic conditions. The grain size distributions (GSD) of the deformed samples were quantified using the D&gt; value (slope of log(frequency) ‐log(radius) of the GSD) for quartz and feldspar fault gouge. Healing causes a decrease in the D&gt; value from &gt;2.0 to ∼1.5. The time dependence of the D&gt; decrease is described by a hydrostatic healing law of the form ΔD = D&gt;(t) − Df = A · e(−λ·t). The results of the laboratory experiments were compared to three natural fault systems, (1) Nojima Fault Zone (Japan), (2) fault zones in the Black Forest (Germany), and (3) Orobic Thrust (Italian Alps). Natural and experimental gouges have similar D&gt; values. Healing is only observed in monomineralic aggregates; polymineralic (i.e., mixed) fault gouges retain their high D&gt; value after extended healing times because grain growth is inhibited. Healing under nonhydrostatic conditions is more rapid than hydrostatic healing. The low strain rates, which were measured during nonhydrostatic healing, are temperature‐dependent and suggest that diffusive mass transfer processes take place during deformation. Thus, fault rocks at upper to midcrustal depth may deform by combined cataclasis and diffusive mass transfer.

Attempt to determine U-Th ages of calcite veins in the Nojima fault zone, Japan

Fluids within fault zones can change the physical and chemical properties of fault materials and have a substantial influence on faulting. Therefore it is important to understand the nature and timing of fluid migration through fault zones. In order to evaluate the timing of fluid infiltration into a deep fault zone on Awaji Island in southern Japan, we attempted to measure U-Th radioactive disequilibrium of carbonate minerals from the Nojima fault, which triggered the 1995 Kobe earthquake. Bulk carbonate samples of vein material have U concentrations of 1.2 to 2 ppm and Th concentrations of 1.6 to 6.5 ppm. The high Th abundances derived from detritus requires correction for initial 230Th in order to acquire an accurate U-Th age. Bulk carbonate samples plot near the equiline in a (230Th/232Th)-(234U/232Th) diagram, suggesting that they precipitated around or before 450 ka. Monte Carlo simulation with finite-age constraint in the IsoPlot program yields a 230Th/234U age of 486 (+380, -190) ka. The presence of radioactive disequilibrium in (234U/238U) indicates that the calcite precipitation age was younger than 1 Ma. The results of leaching experiments indicate that a leaching procedure after firing causes considerable laboratory fractionation of U and Th and does not provide usable age information.

Internal structure and permeability of the Nojima fault, southwest Japan

We conducted permeability measurements on representative fault rocks and surrounding country rocks taken from the Nojima fault zone, which was activated during the 1995 Kobe earthquake, under isotropic confining pressures up to 180 MPa. The results show that the Nojima fault zone consists of a low-permeability fault gouge zone (10 −20–10 −19 m 2 at 180 MPa) within a high-permeability damaged zone of fault breccia and fractured host rock (10 −18–10 −14 m 2 at 180 MPa). The fault gouge zone acts as a barrier to fluid flow across the fault, whereas the surrounding damage zone acts as a fluid conduit. The nature of this proposed permeability structure is consistent with the results of tests conducted on drillcore samples collected from the Nojima fault at depths of 0.6 and 1.8 km. We therefore propose that the permeability of a fault measured from fault rocks exposed at the surface can be used as a representative value for the fault to depths of up to 2 km. We also examined the possibility that thermal pressurization occurred upon the Nojima fault during the Kobe earthquake, based on the obtained permeability data. We found that frictional heating during the Kobe earthquake would have led to an increase in pore pressure at depths below 4 km, thereby resulting in a marked reduction in frictional resistance upon the fault.

Architectural evolution of the Nojima fault and identification of the activated slip layer by Kobe earthquake

Evolutionary history of Nojima Fault zone is clarified by comprehensive examinations of petrological, geophysical, and geochemical characterizations on a fault zone in deep‐drilled core penetrating the Nojima Fault. On the basis of the results, we reconstruct a whole depth profile of the architecture of the Nojima Fault and identify the primal slip layer activated by 1995 Kobe earthquake. The deepest part (8‐ to 12‐km depth) of the fault zone is composed of thin slip layers of pseudotachylite (5 to 10 mm thick each, 10 cm in total). Middle depth (4‐ to 8‐km depth) of the fault zone is composed of fault core (6 to 10 m thick), surrounded by thick (100 m thick) damage zone, characterized by zeolite precipitation. The shallow part of the fault zone (1‐ to 4‐km depth) is composed of distributed narrow shear zones, which are characterized by combination of thin (0.5 cm thick each, 10 cm in total) ultracataclasite layers at the core of shear zones, surrounded by thicker (1 to 3 m thick) damage zones associated with carbonate precipitation. An extremely thin ultracataclasite layer (7 mm thick), activated by the 1995 Kobe earthquake, is clearly identified from numerous past slip layers, overprinting one of the shear zones, as evidenced by conspicuous geological and geophysical anomalies. The Nojima Fault zone was 10 to 100 times thicker at middle depth than that of shallower and deeper depths. The thickening would be explained as a combination of physical and chemical effects as follows. (1) Thickening of “fault core” at middle depth would be attributed to normal stress dependence on thickness of the shear zone and (2) an extreme thickening of “damage zone” in middle depth of the crust would result from the weakening of the fault zone due to super hydrostatic fluid pressure at middle depths. The high fluid pressure would result from faster sealing with low‐temperature carbonate at the shallower fault zone.

Grain size distributions of fault rocks: A comparison between experimentally and naturally deformed granitoids

We have investigated the grain size distribution (GSD) of granitoid fault rock by comparing experimentally produced gouge with fault rock from the Nojima Fault Zone. Triaxial experiments were carried out on wet and dry intact samples of Verzasca Gneiss at T = 300 and 500 °C, P c = 500 and 1030 MPa, ε ˙ = 0.013 – 1.6 × 10 − 4 s − 1 . The GSD has been determined from SEM-micrographs and is characterized by the slope, D, of its log(frequency)–log(radius) histogram. The GSD is not fractal; we observe two slopes for all GSDs. The larger grains in experimentally deformed samples have a D-value, D >, of 2.04 and 2.26 for feldspar and quartz gouge. Cracked grains yield values of D = 1.5–1.6. Increasing the confining pressure or temperature decreases the D-value. For grains smaller than ∼2 μm another D-value, D <, of 0.9–1.1 is observed. The grain radius at the slope change, r K, corresponds to the grinding limit of quartz, so that r K probably represents a change in the dominant comminution mechanism from grinding to attrition processes. The GSD obtained from deformation experiments agrees well with results for the Nojima Fault Zone: D > = 2.02 for gouge and 1.64 for cracked grains; D < = 0.97. Grain size reduction in fault zones develops by a two-stage process: rupturing creates cracked grains; further displacement of fragments causes further comminution by wear and attrition. Cracked grains have been used to calculate the surface energy associated with faulting; it follows that this energy forms a small fraction in the total energy-budget of earthquakes.

Time evolution of hydraulic and electrokinetic parameters around the Nojima fault, Japan, estimated by an electrokinetic method

The Nojima Fault Zone Probe was designed to study the properties and healing processes of the Nojima fault, which is the surface fault rupture of the Hyogo-ken Nanbu earthquake ( M7.2) of 1995 (1995 Kobe earthquake). In this project, water injection experiments were conducted in a borehole of 1800 m depth at the Nojima fault. We set up electrodes around the borehole and observed self-potential variations to investigate the magnitude of electrokinetic and hydraulic parameters around the Nojima fault zone. In the 1997 experiment, self-potential variations were in the range of a few to about 20 mV across 320–450 m electrode dipoles with hydraulic pressure variations from 3.5 to 4 MPa. In the 2000 experiment, self-potential variations were in the range of a few to about 85 mV across 160–260 m electrode dipoles with the hydraulic pressure variations from 3 to 4.5 MPa. In the 2003 experiment, self-potential variations were in the range of a few to about 30 mV across 20–80 m electrode dipoles with hydraulic pressure of 4 MPa. These observed self-potential variations were explained well with an electrokinetic effect due to the underground flow of the injected water. From the observed results, we estimated that the ratio of hydraulic parameters (permeability, porosity, and tortuosity) to electrokinetic parameters (zeta potential and dielectric constant) decreased approximately 40% during eight years after the earthquake. This result suggests that the healing process around the fault zone progress.

Borehole water and hydrologic model around the Nojima fault, SW Japan

The active fault drilling at Nojima Hirabayashi after the 1995 Hyogoken-nanbu (Kobe) earthquake (M JMA = 7.2) provides us with a unique opportunity to investigate subsurface fault structure and the in-situ properties of fault and fluid. The borehole intersected the fault gouge of the Nojima fault at a depth interval of 623 m to 625 m. The lithology is mostly Cretaceous granodiorite with some porphyry dikes. The fault core is highly permeable due to fracturing. The borehole water was sampled in 1996 and 2000 from the depth interval between 630 and 650 m, just below the fault core. The chemical and isotopic compositions were analyzed. Carbon and oxygen isotope ratios of carbonates from the fault core were analyzed to estimate the origin of fluid. The following conclusions were obtained. (1) The ionic and isotopic compositions of borehole water did not change from 1996 to 2000. They are mostly derived from local ground water as mentioned by Sato and Takahashi [Sato, T., Takahashi, M., 2000. Chemical and isotopic compositions of groundwater obtained from the Hirabayashi well. Geological Survey of Japan Interim Report No. EQ/00/1, 187–192.]. (2) Geochemical speciation revealed that the borehole water was derived from a relatively deep reservoir, which may be situated at a depth of 3 to 4 km where the temperature is about 80–90 °C. (3) The shallower part of the Nojima fault (shallower than the reservoir depth) has not been healed from the hydrological viewpoints 5 years after the event, in contrast to the rapid healing detected by S wave splitting [Tadokoro, K., Ando, M., 2002. Evidence for rapid fault healing derived from temporal changes in S wave splitting, Geophys. Res. Lett., 29, 10.1029/2001GL013644.]. (4) Precipitation of calcite from the present borehole water since drilling supports the idea of precipitation of some calcite in coseismic hydraulic fractures in the fault core [Boullier, A-M., Fujimoto, K., Ohtani, T., Roman-Ross, G., Lewin, E., Ito, H., Pezard, P., Ildefonse, B., 2004. Textural evidence for recent co-seismic circulation of fluids in the Nojima fault zone, Awaji Island, Japan., Tectonophysics, 378, 165–181.]. (5) Carbon and oxygen isotope ratios of calcite indicated that the meteoric water flux had been localized at the fault core. (6) A difference in the carbon isotope ratio between the footwall and the hanging wall suggests that the fault has been acted as a hydrologic barrier, although the permeability along the fault is still high.

Tectonic implications of damage zone-related fault-fracture networks revealed in drill core through the Nojima fault, Japan

1800 m of drill core through the Nojima fault zone, Japan, reveals subsidiary fault and fracture networks that developed in the fault zone that triggered the 1995 Ms 7.2 Kobe earthquake. The subsidiary fault zones contain a fault gouge of < 1 cm bounded by thin zones of foliated cataclasite or breccia. Fractures are filled with calcite veins, calcite-cemented breccias, clay, and iron-oxide and carbonate alternation of the granitic host rock. These features are typical of extensional fractures that form the conduit network for fluid flux close to a major fault zone. The zone of distributed deformation surrounding the main fault is 50 m in width, and the dip of the Nojima fault at > 1 km depth is 75°. The fault-fracture networks associated with the Nojima fault zone are coseismic and were filled with carbonate and fine-grained material during repeated seismic-related infiltration of the fault zone by carbonate-bearing subsurface water. This study shows that fault-related fracture networks plays an important role as fluid flow conduits within seismically active faults, and can change in character from zones of high permeability to low permeability due to cementation and/or pore collapse.

Temporal change in permeability of the Nojima fault zone by repeated water injection experiments

Three boreholes were drilled near the Nojima fault, which the 1995 Hyogoken–Nanbu earthquake occurred on. In order to research the properties and the healing process of the fault, water injection experiments were conducted every 3 years. In this report, we researched the permeability of the fault as a measurement of crack density or porosity of the fault zone. Pore water pressure changes in rock due to the water injections at one borehole were observed as discharge changes or groundwater level changes at the other borehole. Using numerical calculations, the permeability of the fault fracture zone was estimated for each experiment. The permeability has been decreasing as time passed, which is thought to show the fault healing process of the Nojima fault after the 1995 Hyogoken–Nanbu earthquake.

Nojima and MTL fault zone gouge dating, Japan

Nojima and MTL fault zone gouge dating, Japan

Correction to “Induced earthquakes accompanying the water injection experiment at the Nojima fault zone, Japan: Seismicity and its migration”

Correction to “Induced earthquakes accompanying the water injection experiment at the Nojima fault zone, Japan: Seismicity and its migration”

Zircon Fission-Track Thermochronology and Applications to Fault Studies

Zircon is one of the most useful minerals used to unravel the Earth’s history recorded in rocks. After the success of etching fission tracks in zircon (Fleischer et al. 1964; Naeser 1969; Krishnaswami et al. 1974), early studies focused primarily on dating young volcanic horizons for stratigraphic purposes (e.g., Naeser et al. 1973; Hurford et al. 1976; Seward 1979; Gleadow 1980; Kowallis et al. 1986; Kohn et al. 1992). Subsequently, fission-track analysis of zircon has been extensively employed, along with other radiometric dating methods, such as U/Pb and (U-Th)/He techniques (e.g., Davis et al. 2003; Reiners 2005), to understand the thermochronology of rocks in a variety of geological settings: i.e., thermal history analysis of basement rocks for orogenic studies (e.g., Zeilter et al. 1982; Hurford 1986; Kamp et al. 1989; Sorkhabi 1993; Seward and Mancktelow 1994; Hasebe and Tagami 2001; Spikings et al. 2001), thermochronology using detrital zircon grains in sedimentary rocks for provenance analysis and thermal history analysis in or near faults, which will be highlighted below. In this present chapter, the thermal sensitivity of the zircon fission-track thermochronometry is reviewed in the context of laboratory and field studies. Then, laboratory procedures of the zircon fission-track analysis using the external detector method are outlined, as practiced in the fission-track laboratory of Kyoto University. For basic fundamentals of these fission-track analyses, including data analysis and graphical displays, see the chapter by Tagami and O’Sullivan (2005). Finally, as an example of recent geological applications of the technique, the results of zircon fission-track analysis of the Nojima fault zone, Japan, are presented. These rocks were the target of systematic drilling of an active fault system using five boreholes and a trench. Applications to other geological settings and the …

Textural evidence for recent co-seismic circulation of fluids in the Nojima fault zone, Awaji island, Japan

The Hirabayashi borehole (Awaji Island, Japan) was drilled by the Geological Survey of Japan (GSJ) 1 year after the Hyogo-ken Nanbu (Kobe) earthquake (1995, M JMA=7.2). This has enabled scientists to study the complete sequence of deformation across the active Nojima fault, from undeformed granodiorite to the fault core. In the fault core, different types of gouge and fractures have been observed and can be interpreted in terms of a complex history of faulting and fluid circulation. Above the fault core and within the hanging wall, compacted cataclasites and gouge are cut by fractures which show high apparent porosity and are filled by 5–50 μm euhedral and zoned siderite and ankerite crystals. These carbonate-filled fractures have been observed within a 5.5-m-wide zone above the fault, but are especially abundant in the vicinity (1 m) of the fault. The log-normal crystal size distributions of the siderite and ankerite suggest that they originated by decaying-rate nucleation accompanied by surface-controlled growth in a fluid saturated with respect to these carbonates. These carbonate-filled fractures are interpreted as the result of co-seismic hydraulic fracturing and upward circulation of fluids in the hanging wall of the fault, with the fast nucleation of carbonates attributed to a sudden fluid or CO 2 partial pressure drop due to fracturing. The fractures cut almost all visible structures at a thin section scale, although in some places, the original idiomorphic shape of carbonates is modified by a pressure-solution mechanism or the carbonate-filled fractures are cut and brecciated by very thin gouge zones; these features are attributed to low and high strain-rate mechanisms, respectively. The composition of the present-day groundwater is at near equilibrium or slightly oversaturated with respect to the siderite, calcite, dolomite and rhodochrosite. Taken together, this suggests that these fractures formed very late in the evolution of the fault zone, and may be induced by co-seismic hydraulic fracturing and circulation of a fluid with a similar composition to the present-day groundwater. They are therefore potentially related to recent earthquake activity (<1.2 Ma) on the Nojima fault.