Richard H. Sibson

Fault rocks and fault mechanisms

Physical factors likely to affect the genesis of the various fault rocks—frictional properties, temperature, effective stress normal to the fault and differential stress—are examined in relation to the energy budget of fault zones, the main velocity modes of faulting and the type of faulting, whether thrust, wrench, or normal. In a conceptual model of a major fault zone cutting crystalline quartzo-feldspathic crust, a zone of elastico-frictional (EF) behaviour generating random-fabric fault rocks (gouge—breccia—cataclasite series—pseudotachylyte) overlies a region where quasi-plastic (QP) processes of rock deformation operate in ductile shear zones with the production of mylonite series rocks possessing strong tectonite fabrics. In some cases, fault rocks developed by transient seismic faulting can be distinguished from those generated by slow aseismic shear. Random-fabric fault rocks may form as a result of seismic faulting within the ductile shear zones from time to time, but tend to be obliterated by continued shearing. Resistance to shear within the fault zone reaches a peak value (greatest for thrusts and least for normal faults) around the EF/QP transition level, which for normal geothermal gradients and an adequate supply of water, occurs at depths of 10–15 km.

Structural permeability of fluid-driven fault-fracture meshes

Fluid redistribution in the crust is influenced by hydraulic gradient, by existing permeability anisotropy arising from bedding and other forms of layering, and by structural permeability developed under the prevailing stress field. Field evidence suggests that mesh structures, comprising faults interlinked with extensional-shear and purely extensional vein-fractures, form important conduits for large volume flow of hydrothermal and hydrocarbon fluids. Meshes may be ‘self-generated’ by the infiltration of pressurised fluids into a stressed heterogeneous rock mass with varying material properties, developing best where bulk coaxial strain is symmetric with existing layering, but they also form under predominantly simple shear. Fluid passage through such structures generates earthquake swarm activity by distributed fault-valve action along suprahydrostatic gradients that may arise from compaction overpressuring, metamorphic dewatering, magmatic intrusion, and mantle degassing. Within mesh structures, strong directional permeability may develop in the σ2 direction parallel to fault-fracture intersections and orthogonal to fault slip vectors. In particular tectonic settings, this promotes strongly focused flow with high potential for mineralisation. Mesh activation requires the condition Pf ~ σ3 to be maintained for the structures to remain high permeability conduits, requiring fluid overpressuring at other than shallow depths in extensional-transtensional regimes. Favoured localities for mesh development include linkage structures along large-displacement fault zones such as dilational jogs, lateral ramps, and transfer faults. In some circumstances, mesh formation appears to precede the development of major faults.

Implications of fault-valve behaviour for rupture nucleation and recurrence

Sibson, R.H., 1992. Implications of fault-valve behaviour for rupture nucleation and recurrence. In: T. Mikumo, K. Aki, M. Ohnaka, L.J. Ruff and P.K.P. Spudich (Editors), Earthquake Source Physics and Earthquake Precursors. Tectonophyslcs, 211: 283–293. Earthquakes in the shallow crust are generally thought to arise from the frictional instability of existing faults under shear stress. In simple recurrence models, failure on such faults is assumed to occur when tectonic shear stress (τ) rises to some constant critical level. However, frictional strength (τf) may also vary substantially through the interseismic period, in which case recurrence intervals between successive earthquakes are related to the time-dependence of τf as well as τ . In part, the manner in which τf changes with τ depends on the coupling between normal stress and shear stress on the fault, and is related to the mode of faulting and the manner of fault loading. Time-dependence of cohesive strength and static friction may also play a role. However, fluid pressure cycling as a consequence of fault-valve behaviour (where the fault transects a suprahydrostatic gradient in fluid pressure) may give rise to very large variations in fault strength. Geological evidence suggests that value-action may be especially important in the lower regions of the seismogenic zone, where large ruptures tend to nucleate, the largest fluid pressure fluctuations being associated with faults that remain active as a consequence of fluid overpressure though severely misoriented for reactivation in the prevailing stress field. Extensive hydrofracture dilatancy is likely to develop prefailure in the vicinity of such faults. Changes in frictional strength over the interseismic period as a result of fault-valve activity (Δτf) may greatly exceed the shear stress drop at failure (1 < Δτ < 10 MPa, typically). In such circumstances, recurrence intervals between successive events can be highly variable. A rich variety of recurrence behaviour becomes possible, depending on the relative magnitudes of Δτf and Δτ, and the coupling between shear stress accumulation and changing fluid pressure levels through the interseismic period.

A note on fault reactivation

Reactivation of existing faults whose normal lies in the σ1′σ3′ plane of a stress field with effective principal compressive stresses σ1′ >σ2′ >σ3′ is considered for the simplest frictional failure criterion, τ = μσn′ = μ(σn − P), where τ and σn are respectively the shear and normal stresses to the existing fault, P is the fluid pressure and μ is the static friction. For a plane oriented at θ to σ1′, the stress ratio for reactivation is (σ1′/σ3′) = (1 + μ cot θ)/(1 − μ tan θ). This ratio has a minimum positive value at the optimum angle for reactivation given by θ∗ = 12tan−1 (1/μ) but reaches infinity when θ = 2θ∗, beyond which σ3′ < 0 is a necessary condition for reactivation. An important consequence is that for typical rock friction coefficients, it is unlikely that normal faults will be reactivated as high-angle reverse faults or thrusts as low-angle normal faults, unless the effective least principal stress is tensile.

Continental fault structure and the shallow earthquake source

Plate boundaries in continental crust are generally less sharply defined than in the oceans, with seismicity spread over broad areas. Interplate displacements appear to be largely accommodated by networks of major fault zones. A simple 2-level model for these important structures accounts for the depth distribution of most continental earthquakes, and for the observed range of faulting styles and associated rock deformation textures. The model consists of a seismogenic frictional slip regime overlying quasi-plastic mylonite belts wherein shearing is largely accommodated aseismically, due mainly to the changing response of quartz to deformation with increasing temperature. Shear resistance increases with depth to a peak value in the vicinity of the frictiona1/quasi-plastic transition and then decreases rapidly. The depth to which microseismic activity extends appears inversely related to regional heat flow and can be satisfactorily modelled as the frictional/quasi-plastic transition for different geotherms using laboratory determined flow laws for quartz-bearing rocks. Larger earthquake ruptures (M > 5.5) tend to nucleate near the base of the seismogenic regime in the region inferred to have the highest shear resistance and concentration of distortional strain energy. Consideration is also given to the depression of isotherms and seismic activity in regions of thrusting, and to the question of the downward continuation of major fault zones through the lithosphere. Decoupling of the upper crust on flat-lying shear zones may accompany higher-level dip-slip (and perhaps in some circumstances, strike-slip) faulting, being favoured by above average continental heat flow and a high quartz content in the middle or deep crust. The average level of deviatoric stress within the seismogenic regime remains an outstanding problem.

Earthquake rupturing as a mineralizing agent in hydrothermal systems

Much fault-hosted epithermal mineralization is localized in dilational jogs between en echelon fault segments, as fissure veins or as hydrothermally cemented, high-dilation wall-rock breccias. Jog widths may range from millimetres to kilometres; vein textures record histories of incremental development. Perturbation or arrest of earthquake ruptures at dilational jogs has been observed and is believed to involve extensional fracturing at the rupture tip, locally reducing fluid pressure and inducing suctions opposing rapid slip transfer across the jog. This forced fissuring leads to brecciation by hydraulic implosion and to a concentrated fluid influx, allowing delayed slip transfer accompanied by aftershock activity. Within the southern San Andreas fault system, major dilational jogs extend throughout the seismogenic regime and form loci for magmatic-hydrothermal systems; they act as vertical pipelike conduits for enhanced fluid flow. Rupture termination at these structures has sometimes been followed by hydrothermal eruptions, suggesting that high-level boiling events are triggered by the arrest mechanism. It thus seems probable that episodic mineral deposition in the top 1–2 km of such jogs is induced by the dynamic effects of rupturing on the flanking strike-slip faults.

Introduction to special section: Mechanical involvement of fluids in faulting

A growing body of evidence suggests that fluids are intimately linked to a variety of faulting processes. These include the long term structural and compositional evolution of fault zones; fault creep; and the nucleation, propagation, arrest, and recurrence of earthquake ruptures. Besides the widely recognized physical role of fluid pressures in controlling the strength of crustal fault zones, it is also apparent that fluids can exert mechanical influence through a variety of chemical effects. The United States Geological Survey sponsored a Conference on the Mechanical Effects of Fluids in Faulting under the auspices of the National Earthquake Hazards Reduction Program at Fish Camp, California, from June 6 to 10, 1993. The purpose of the conference was to draw together and to evaluate the disparate evidence for the involvement of fluids in faulting; to establish communication on the importance of fluids in the mechanics of faulting between the different disciplines concerned with fault zone processes; and to help define future critical investigations, experiments, and observational procedures for evaluating the role of fluids in faulting. This conference drew together a diverse group of 45 scientists, with expertise in electrical and magnetic methods, geochemistry, hydrology, ore deposits, rock mechanics, seismology, and structural geology. Some of the outstanding questions addressed at this workshop included the following: 1. What are fluid pressures at different levels within seismically active fault zones? Do they remain hydrostatic throughout the full depth extent of the seismogenic regime, or are they generally superhydrostatic at depths in excess of a few kilometers? 2. Are fluid pressures at depth within fault zones constant through an earthquake cycle, or are they time-dependent? What is the spatial variability in fluid pressures? 3. What is the role of crustal fluids in the overall process of stress accumulation, release, and transfer during the earthquake cycle? Through what mechanisms might fluid pressure act to control the processes of rupture nucleation, propagation, and arrest? 4. What is the chemical role of fluids in facilitating fault creep, including their role in aiding solid-state creep and brittle fracture processes and in facilitating solution-transport deformation mechanisms? 5. What are the chemical effects of aqueous fluids on constitutive response, fractional stability, and long-term fault strength? 6. What are the compositions and physical properties of faultfluids at different crustal levels? 7. What are the mechanisms by which porosity and permeability are either created or destroyed in the middle to lower crust? What factors control the rates of these processes? How should these effects be incorporated into models of time-dependent fluid transport in fault zones? 8. What roles do faults play in distributing fluids in the crust and in altering pressure domains? In other words, when and by what mechanisms do faults aid in or inhibit fluid migration? What are the typical fluid/rock ratios, flow rates, and discharges for fault zones acting as fluid conduits? 9. Are fluids present in the subseismogenic crust, and by what transformation and/or transport processes are they incorporated into the shallower seismogenic portions of faults?

Earthquake faulting as a structural process

Abstract Structural geology is concerned with the history of movement in the Earths crust and the processes by which displacements occur. In the upper one third to one half of deforming continental crust, displacement is accommodated largely by seismic slip increments on existing faults. It follows that earthquakes and related processes are an integral part of structural geology. Traditionally, structural geologists have been preoccupied with the complexity of the finite deformation within fault zones and with the stress states prevailing at the initiation of faults in intact crust. Future structural work should be directed more towards understanding the dynamic character of fault reactivation during incremental slip, and related effects. Questions of interest include rheological and geometrical controls on the initiation, perturbation and termination of ruptures; directivity effects associated with rupture propagation; the recognition of structures resulting from repeated stress cycling within seismogenic crust; and identification of structural features diagnostic of shear stress levels during faulting. Structures arising from the inter-relationships between slip episodes and induced fluid flow are of special importance, because these dynamic fault processes appear influential in the development of much fault-hosted mineralization. Mesothermal gold-quartz lodes hosted in high-angle reverse shear zones of mixed brittle-ductile character form illustrative examples of structures that, arguably, can only be interpreted by seismo-structural analysis embodying the concepts listed above.

Thickness of the Seismic Slip Zone

This article reviews geologic and other evidence constraining the thickness of the principal slip zone (PSZ) that accommodates the bulk of coseismic shear displacement during an individual rupture event. Surface deformation from rupturing may occupy swaths tens of meters or more in width, but trenches across active faults generally reveal that incremental slip is accommodated by a PSZ that is tens of centimeters or less in thickness. Geomorphic evidence, coupled with the observations from trenching, suggest a PSZ may stay well localized for distances of several kilometers through many rupture episodes. Mine exposures and exhumed fault zones demonstrate that PSZs separating different lithologies within the “fault core,” although contained within “damage zones” of variably fractured rock ranging up to hundreds of meters in thickness, often comprise just a few centimeters of gouge/ultracataclasite that have accommodated large finite displacements (>1 km). Microstructural studies demonstrate incremental slip localized still further down to 1–10 mm, as do other fault-rock assemblages (slickensides and slickenfibers, fault-veins of pseudotachylyte friction-melt, intravein septa in hydrothermal fault infills). The accumulated evidence indicates that localization of coseismic shearing to less than 10 cm on planar faults is widespread throughout the crustal seismogenic zone, with extreme localization to less than 1 cm not uncommon. However, some distributed coseismic shear may also develop, especially at rupture irregularities. Coseismic reduction of shear resistance from friction-melting (Δ T ∼ 1000°C) or from transient thermal pressurization of aqueous fluids (Δ T ∼ 100°C) requires slip during moderate-to-large earthquakes ( u > 1 m) to be restricted to narrow zones, respectively a few centimeters or tens of centimeters in thickness. Given the evidence for slip localization, the apparent scarcity of pseudotachylyte suggests either that seismic friction-melting is a rare phenomenon, or that pseudotachylyte is only rarely preserved in recognizable form within mature hydrated fault zones.

Fluid involvement in normal faulting

Evidence of fluid interaction with normal faults comes from their varied role as flow barriers or conduits in hydrocarbon basins and as hosting structures for hydrothermal mineralisation, and from fault-rock assemblages in exhumed footwalls of steep active normal faults and metamorphic core complexes. These last suggest involvement of predominantly aqueous fluids over a broad depth range, with implications for fault shear resistance and the mechanics of normal fault reactivation. A general downwards progression in fault rock assemblages (high-level breccia-gouge (often clay-rich) → cataclasites → phyllonites → mylonite → mylonitic gneiss with the onset of greenschist phyllonites occurring near the base of the seismogenic crust) is inferred for normal fault zones developed in quartzo-feldspathic continental crust. Fluid inclusion studies in hydrothermal veining from some footwall assemblages suggest a transition from hydrostatic to suprahydrostatic fluid pressures over the depth range 3–5 km, with some evidence for near-lithostatic to hydrostatic pressure cycling towards the base of the seismogenic zone in the phyllonitic assemblages. Development of fault-fracture meshes through mixed-mode brittle failure in rock-masses with strong competence layering is promoted by low effective stress in the absence of thoroughgoing cohesionless faults that are favourably oriented for reactivation. Meshes may develop around normal faults in the near-surface under hydrostatic fluid pressures to depths determined by rock tensile strength, and at greater depths in overpressured portions of normal fault zones and at stress heterogeneities, especially dilational jogs. Overpressures localised within developing normal fault zones also determine the extent to which they may reutilise existing discontinuities (for example, low-angle thrust faults). Brittle failure mode plots demonstrate that reactivation of existing low-angle faults under vertical σ1 trajectories is only likely if fluid overpressures are localised within the fault zone and the surrounding rock retains significant tensile strength. Migrating pore fluids interact both statically and dynamically with normal faults. Static effects include consideration of the relative permeability of the faults with respect to the country rock, and juxtaposition effects which determine whether a fault is transmissive to flow or acts as an impermeable barrier. Strong directional permeability is expected in the subhorizontal σ2 direction parallel to intersections between minor faults, extension fractures, and stylolites. Three dynamic mechanisms tied to the seismic stress cycle may contribute to fluid redistribution: (i) cycling of mean stress coupled to shear stress, sometimes leading to postfailure expulsion of fluid from vertical fractures; (ii) suction pump action at dilational fault jogs; and, (iii) fault-valve action when a normal fault transects a seal capping either uniformly overpressured crust or overpressures localised to the immediate vicinity of the fault zone at depth. The combination of σ2 directional permeability with fluid redistribution from mean stress cycling may lead to hydraulic communication along strike, contributing to the protracted earthquake sequences that characterise normal fault systems.