Hydrothermal Friction Experiments on Simulated Basaltic Fault Gouge and Implications for Megathrust Earthquakes

Abstract

Megathrust earthquakes in subduction zones generally source in the depth range of ∼5-25 km corresponding to a temperature (T) range of ∼150-350°C (Hyndman et al., 1997). The extent of this "seismogenic zone" was originally postulated to be related to the smectite-illite dewatering transition as the transition temperature is consistent with the temperature condition at the up-dip limit of the seismogenic zone (Oleskevich et al., 1999). However, this hypothesis is not fully supported by experiments (e.g., Saffer & Marone, 2003). Instead, recent experimental studies showed that the effect of temperature on the velocity dependence of illite-quartz mixture can explain both up-dip and down-dip limits (Den Hartog, Peach, et al., 2012), and that a change in the extent of lithification of sediments during subduction may contribute to making the plate boundary fault (décollement) frictionally unstable (Ikari & Hüpers, 2021).

Microphysical Modeling of Fault Slip and Stability Transition in Hydrothermal Conditions

“…Moreover, the value of (a − b) at 0.1-0.01 μm/s is a bit lower than that at 1-0.1 μm/s, implying that the load-point velocity can influence the friction rate parameter (a − b) and also the temperature range for stability transitions. The velocity-weakening behavior is enhanced by slow load-point velocities, which has been verified in some experiments (Chen et al, 2017;Mei et al, 2022;Niemeijer & Collettini, 2014;Okuda et al, 2023).…”

“…The experimental results provide important constraints on physical models of fault friction by showing how temperature, effective stress, and lithology affect frictional strength and frictional parameters, for example, friction coefficient, direct friction effect, and evolutionary friction effect. In particular, some high‐temperature friction experiments demonstrated the frictional‐plastic transition (i.e., frictional slip to plastic flow), comparable to the brittle‐ductile transition in the crust, as well as the transition of velocity dependence of friction at different temperature ranges, which is consistent with the transition of aseismic‐seismic‐aseismic zones with depth (Blanpied et al., 1995; Den Hartog et al., 2012; Den Hartog & Spiers, 2013; Niemeijer et al., 2016; Okuda et al., 2023). Based on laboratory observations, several constitutive friction relations have been proposed to model fault slip and explain the underlying mechanisms in hydrothermal conditions, such as an empirical rate‐ and temperature‐dependent constitutive law (Blanpied et al., 1995; Chester, 1994), a power‐law temperature‐, rate‐, and state‐dependent friction with multiple healing mechanisms (Barbot, 2022), an empirical friction‐to‐flow relation (Shimamoto & Noda, 2014), and a physics‐based model for steady‐state friction (Aharonov & Scholz, 2019).…”

“…The end of velocity-weakening behaviors occurs at temperatures of 320°C-500°C, implying the depth for seismic-aseismic transition is around 14-22 km. The experimental (Blanpied et al, 1995), compared with illite-quartz gouges (Den Hartog et al, 2012), muscovite-quartz gouges (Den Hartog & Spiers, 2013), gouges from Alpine Fault PSZ (Niemeijer et al, 2016), and altered basalt gouge (Okuda et al, 2023). Temperature-depth relation is from Kolawole et al (2019).…”

“…Extensive laboratory studies have documented the temperature dependence of frictional properties and associated physical processes for a variety of representative rocks in different tectonic faults, such as pure gouges of granite (An et al, 2022;Blanpied et al, 1991Blanpied et al, , 1995Kolawole et al, 2019), quartz (Chester, 1994;Chester & Higgs, 1992), basalt (Okuda et al, 2023), and blueschist (Sawai et al, 2016), mixed gouges of quartz-biotite (Lu & He, 2018), illite-quartz (Den Hartog et al, 2012), muscovite-quartz (Den Hartog & Spiers, 2013), and actinolite-chlorite (Okamoto et al, 2020), as well as some natural gouges from the Alpine Fault (Niemeijer et al, 2016), Nankai subduction zone (Den Hartog et al, 2022), and Changning Block (An et al, 2020). The experimental results provide important constraints on physical models of fault friction by showing how temperature, effective stress, and lithology affect frictional strength and frictional parameters, for example, friction coefficient, direct friction effect, and evolutionary friction effect.…”

“…Figure 4. Depth profile of (a − b) values for granite gouges at the velocities of 0.01-0.1-1 μm/s(Blanpied et al, 1995), compared with illite-quartz gouges(Den Hartog et al, 2012), muscovite-quartz gouges(Den Hartog & Spiers, 2013), gouges from Alpine Fault PSZ(Niemeijer et al, 2016), and altered basalt gouge(Okuda et al, 2023). Temperature-depth relation is fromKolawole et al (2019).…”