Takehiro Hirose

Fault lubrication during earthquakes

The determination of rock friction at seismic slip rates (about 1 m s−1) is of paramount importance in earthquake mechanics, as fault friction controls the stress drop, the mechanical work and the frictional heat generated during slip. Given the difficulty in determining friction by seismological methods, elucidating constraints are derived from experimental studies. Here we review a large set of published and unpublished experiments (∼300) performed in rotary shear apparatus at slip rates of 0.1–2.6 m s−1. The experiments indicate a significant decrease in friction (of up to one order of magnitude), which we term fault lubrication, both for cohesive (silicate-built, quartz-built and carbonate-built) rocks and non-cohesive rocks (clay-rich, anhydrite, gypsum and dolomite gouges) typical of crustal seismogenic sources. The available mechanical work and the associated temperature rise in the slipping zone trigger a number of physicochemical processes (gelification, decarbonation and dehydration reactions, melting and so on) whose products are responsible for fault lubrication. The similarity between (1) experimental and natural fault products and (2) mechanical work measures resulting from these laboratory experiments and seismological estimates suggests that it is reasonable to extrapolate experimental data to conditions typical of earthquake nucleation depths (7–15 km). It seems that faults are lubricated during earthquakes, irrespective of the fault rock composition and of the specific weakening mechanism involved.

Growth of molten zone as a mechanism of slip weakening of simulated faults in gabbro during frictional melting

[1] To understand how frictional melting affects fault instability, we performed a series of high-velocity friction experiments on gabbro at slip rates of 0.85-1.49 m s -1 , at normal stresses of 1.2-2.4 MPa and with displacements up to 124 m. Experiments have revealed two stages of slip weakening; one following the initial slip and the other immediately after the second peak friction. The first weakening is associated with flash heating, and the second weakening is due to the formation and growth of a molten layer along a simulated fault. The two stages of weakening are separated by a marked strengthening regime in which melt patches grow into a thin, continuous molten layer at the second peak friction. The frictional coefficient decays exponentially from 0.8-1.1 to 0.6 during the second weakening. The host rocks are separated completely by a molten layer during this weakening so that the shear resistance is determined by the gross viscosity and shear strain rate of the molten layer. Melt viscosity increases notably soon after a molten layer forms. However, a fault weakens despite the increase in melt viscosity, and the second weakening is caused by the growth of molten layer resulting in the reduction in shear strain rate of the molten layer. Very thin melt cannot be squeezed out easily from a fault zone so that the rate of melting would be the most critical factor in controlling the slip-weakening distance. Effect of frictional melting on fault motion can be predicted by solving a Stefan problem dealing with moving host rock/molten zone boundaries.

Extreme dynamic weakening of faults during dehydration by coseismic shear heating

The dynamic strength of seismogenic faults has a critical effect on earthquake slip instability and seismic energy release. High velocity friction experiments on simulated faults in serpentinite at earthquake slip conditions show a decrease in friction coefficient from 0.6 to 0.15 as the slip velocity reaches 1.1 m/s at normal stresses up to 24.5 MPa. The extraordinary reduction in fault strength is attributed to flash heating at asperity contacts of gouge particles formed during sliding. The rapid heating at asperities causes serpentine dehydration. In impermeable fault zones in nature, serpentine dehydration and subsequent fluid pressurization due to coseismic frictional heating may promote further weakening. This dynamic fault-weakening mechanism may explain the lack of pronounced heat flow in major crustal faults such as the San Andreas.

Fault lubrication and earthquake propagation in thermally unstable rocks

Experiments performed on dolomite or Mg-calcite gouges at seismic slip rates ( v > 1 m/s) and displacements (d > 1 m) show that the frictional coefficient μ decays exponentially from peak values (m p ≈ 0.8, in the Byerlee9s range), to extremely low steady-state values (μ ss ≈ 0.1), attained over a weakening distance D w . Microstructural observations show that discontinuous patches of nanoparticles of dolomite and its decomposition products (periclase and lime or portlandite) were produced in the slip zone during the transient stage (d w ). These observations, integrated with CO 2 emissions data recorded during the experiments, suggest that particle interaction in the slip zone produces flash temperatures that are large enough to activate chemical and physical processes, e.g., decarbonation reactions ( T = 550 °C). During steady state (d ≥ D w ), shear strength is very low and not dependent upon normal stresses, suggesting that pressurized fluids (CO 2 ) may have been temporarily trapped within the slip zone. At this stage a continuous layer of nanoparticles is developed in the slip zone. For d >> D w , a slight but abrupt increase in shear strength is observed and interpreted as due to fluids escaping the slip zone. At this stage, dynamic weakening appears to be controlled by velocity dependent properties of nanoparticles developed in the slip zone. Experimentally derived seismic source parameter W b (i.e., breakdown work, the energy that controls the dynamics of a propagating fracture) (1) matches W b values obtained from seismological data of the A.D. 1997 M6 Colfiorito (Italy) earthquakes, which nucleated in the same type of rocks tested in this study, and (2) suggests similar earthquake-scaling relationships, as inferred from existing seismological data sets. We conclude that dynamic weakening of experimental faults is controlled by multiple slip weakening mechanisms, which are activated or inhibited by physicochemical reactions in the slip zone.

Granular nanoparticles lubricate faults during seismic slip

Nanoparticles are known to form in narrow slip zones in natural and experimental fault zones, but their possible role in dynamic weakening of faults during seismic slip remains almost unexplored. We conducted friction experiments on periclase (MgO) nanoparticles (50 nm in average size) at rates as high as 1.3 m s −1 , a typical speed of seismic slip. The nanoparticles were used as the initial gouge to avoid complexities arising from comminution and fluid release. The dynamic friction decreased with increasing slip rate and the steady-state frictional coefficient reduced to as low as 0.1 at a slip rate of 1.3 m s −1 . Flash heating is not effective for nanoparticles, and we propose that development of slickensides and dominant operation of nanoparticle rolling cause such a weakening. Nanoparticle lubrication appears to be as effective as melt lubrication and thermal pressurization, and the formation of nanoparticles in slip zones may be an important fault lubrication process.

Stress State in the Largest Displacement Area of the 2011 Tohoku-Oki Earthquake

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. Borehole stress measurements indicate a nearly total stress drop in the region of largest slip. The 2011 moment magnitude 9.0 Tohoku-Oki earthquake produced a maximum coseismic slip of more than 50 meters near the Japan trench, which could result in a completely reduced stress state in the region. We tested this hypothesis by determining the in situ stress state of the frontal prism from boreholes drilled by the Integrated Ocean Drilling Program approximately 1 year after the earthquake and by inferring the pre-earthquake stress state. On the basis of the horizontal stress orientations and magnitudes estimated from borehole breakouts and the increase in coseismic displacement during propagation of the rupture to the trench axis, in situ horizontal stress decreased during the earthquake. The stress change suggests an active slip of the frontal plate interface, which is consistent with coseismic fault weakening and a nearly total stress drop.

Fractal dimension of molten surfaces as a possible parameter to infer the slip-weakening distance of faults from natural pseudotachylytes

High-velocity friction experiments on gabbro and monzodiorite, using a rotary-shear high-velocity friction apparatus, have revealed that frictional melting and progressive growth of a molten layer along a fault cause slip weakening, eventually reaching a nearly steady-state. The melting surface at the host rock/molten layer interface is initially very flat, but it becomes more complex and rounded in shape towards the steady state owing to the selective melting of minerals with lower melting points and the Gibbs–Thomson effect. This change in the melting-surface topography can be quantitatively expressed by the fractal dimension D, as determined by the divider method, from about 1.0 near the peak friction to around 1.1 near the steady-state friction. The ultimate fractal dimension at steady-state friction tends to decrease with increasing heat production rate presumably due to more rapid and uniform melting. A systematic correlation of D with mechanical behavior of the fault during frictional melting may provide a way of estimating slip-weakening distance and heat production rate at steady-state friction by measuring D for natural pseudotachylytes on slip surfaces with different displacements. The weakening distance is of vital significance in relation to fault instability and the heat production rate is related to the fault strength. The experimental studies point to ways to estimate these difficult quantities for natural faults.

Drilling constraints on lithospheric accretion and evolution at Atlantis Massif, Mid‐Atlantic Ridge 30°N

Expeditions 304 and 305 of the Integrated Ocean Drilling Program cored and logged a 1.4 km section of the domal core of Atlantis Massif. Postdrilling research results summarized here constrain the structure and lithology of the Central Dome of this oceanic core complex. The dominantly gabbroic sequence recovered contrasts with predrilling predictions; application of the ground truth in subsequent geophysical processing has produced self-consistent models for the Central Dome. The presence of many thin interfingered petrologic units indicates that the intrusions forming the domal core were emplaced over a minimum of 100-220 kyr, and not as a single magma pulse. Isotopic and mineralogical alteration is intense in the upper 100 m but decreases in intensity with depth. Below 800 m, alteration is restricted to narrow zones surrounding faults, veins, igneous contacts, and to an interval of locally intense serpentinization in olivine-rich troctolite. Hydration of the lithosphere occurred over the complete range of temperature conditions from granulite to zeolite facies, but was predominantly in the amphibolite and greenschist range. Deformation of the sequence was remarkably localized, despite paleomagnetic indications that the dome has undergone at least 45 degrees rotation, presumably during unroofing via detachment faulting. Both the deformation pattern and the lithology contrast with what is known from seafloor studies on the adjacent Southern Ridge of the massif. There, the detachment capping the domal core deformed a 100 m thick zone and serpentinized peridotite comprises similar to 70% of recovered samples. We develop a working model of the evolution of Atlantis Massif over the past 2 Myr, outlining several stages that could explain the observed similarities and differences between the Central Dome and the Southern Ridge.

Pelagic smectite as an important factor in tsunamigenic slip along the Japan Trench

The very large slip on the shallow portion of the subduction interface during the 2011 Tohoku-oki earthquake (M w 9.0) caused a huge tsunami along the northeast coast of Honshu, Japan. In order to elucidate the mechanics of such tsunamigenic slip, the Integrated Ocean Drilling Program Expedition 343 (Japan Trench Fast Drilling Project, JFAST), was carried out one year after the earthquake and succeeded in recovering rocks constituting the active plate boundary fault. Our mineralogical analyses using X-ray diffraction reveal that the shallow portion of the fault zone that caused the earthquake is significantly enriched in smectite compared to the surrounding sediments, which may be intimately linked to the tsunamigenic shallow faulting. For comparison, we also analyzed mineralogical features of incoming sediments just prior to subduction, recovered on the outer rise of the Japan Trench (Site 436, Deep Sea Drilling Project Leg 56), and found a characteristic smectite-rich horizon in the uppermost ∼5 m of the pelagic clay layer. This horizon should be mechanically weak and will become the future plate boundary fault, as observed in the JFAST cores. The smectite-rich deposits are broadly distributed in the northwestern Pacific Ocean, and may therefore potentially enhance conditions for large shallow slip during earthquakes that occur over a broad area of the Japan Trench plate boundary, which would result in large tsunamis for this region.

Slip-Weakening Distance of Faults during Frictional Melting as Inferred from Experimental and Natural Pseudotachylytes

The slip-weakening distance, D c , over which a fault weakens during seismogenic fault motion, is one of most important parameters controlling fault instability. A paradox over the last decade has been the source of the large difference between laboratory-measured values of D c using conventional friction experiments (10 −5 –10 −3 m) and seismically determined D c values for natural earthquakes (10 −1 –10 0 m). D c is a difficult quantity to estimate for natural faults, but this study attempts to determine the order of D c values for natural faults containing pseudotachylytes based on a comparison between experimental and natural pseudotachylytes. Experiments at a constant slip velocity of 0.85 m/sec and a normal stress of 1.4 MPa on gabbro have revealed that the progressive growth of a molten layer causes marked slip weakening. Slip weakening (a reduction in frictional strength with displacement) and the growth of the molten layer (an increase in the thickness of the molten layer with displacement) are both characterized by exponential changes with nearly the same characteristic distances. The comparison suggests that the order of D c can be estimated from the relationship between pseudotachylyte thickness along its generation fault and fault displacement. Sibson’s (1975) data from the Outer Hebrides exhibit a similar pseudotachylyte-thickness-versus-fault-displacement curve with a characteristic distance of 0.38 m. This is the same order of magnitude as seismically inferred values of D c . The effect of frictional melting is a possible solution for the D c paradox, although it is not the only possible solution, since frictional melting along large faults is rather uncommon.