Tomowo Hirasawa

A numerical study on seismic coupling along subduction zones using a laboratory-derived friction law

Abstract A numerical simulation study is performed to examine the effect of frictional characteristics of the boundary between a continental plate and a subducting oceanic plate on the seismic coupling between them. We consider a dip-slip fault in a 2-D uniform elastic half-space. The frictional force following a rate- and state-dependent friction law is assumed to act on the fault plane or the plate boundary, on which the values of friction parameters are varied with depth. Simulation results show that great earthquakes repeatedly occur at a shallower part of the plate boundary and stable sliding occurs at a deeper part. Seismic coupling is quantitatively discussed in terms of the seismic coupling coefficient defined by the seismic slip divided by the total amount of the seismic and aseismic slip. The critical fault length required for the occurrence of unstable slip is found to be proportional to the characteristic slip distance and inversely proportional to the coefficient of slip-rate dependence of the steady-state friction. The seismic coupling coefficient is relatively larger (0.6 to 0.8) when the length of the seismogenic zone is much longer than the critical fault length, while the seismic coupling coefficient becomes zero when the seismogenic zone length is close to or shorter than the critical fault length. In the latter case, slow or silent earthquakes are expected to occur. It is further found that spatially non-uniform distributions of friction parameters tend to weaken the seismic coupling.

Strain-rate effect on frictional strength and the slip nucleation process

Abstract Kato, N., Yamamoto, K., Yamamoto, H. and Hirasawa, T., 1992. Strain-rate effect on frictional strength and the slip nucleation process. In: T. Mikumo, K. Aki, M. Ohnaka, L.J. Ruff and P.K.P. Spudich (Editors), Earthquake Source Physics and Earthquake Precursors. Tectonophysics, 211: 269–282. The strain-rate dependence of frictional strength is investigated in relation to sliding behaviour by biaxial compression tests of large-scale granite samples. Frictional sliding is generated by increasing shear strain at a constant rate under a constant normal stress on the fault. The ram displacement is held constant for a prescribed time interval between two successive sliding events so as to have the same time effect of stationary contact between sliding surfaces. We obtained the following results. As the strain rate increases: (1) sliding becomes more unstable; (2) the nucleation zone size of stick-slip decreases; and (3) frictional strength logarithmically increases. Result (3) is consistent with the previous study by Dieterich (1979a), and (1) and (3) are similar to the strain-rate effects on failure characteristics of intact rocks reported in the literature. Result (3) can be explained by a rate- and state-dependent friction law. Our results suggest that the strain-rate dependence of the fault strength plays an important role in controlling the nucleation process of faulting.

Effects of strain rate and strength nonuniformity on the slip nucleation process: A numerical experiment

Abstract We perform a numerical simulation study of the slip nucleation process prior to unstable slip events on pre-existing faults to compare the numerical results with the experimental observations in a laboratory. Our previous experiment of the strain-rate effects on the slip nucleation process indicated that: (1) the macroscopic frictional strength logarithmically increases with the strain rate; (2) the critical length of the slip nucleation zone decreases with an increase in the strain rate; and (3) the rupture propagation speed during the nucleation process increases with an increase in the rupture propagation distance or in the strain rate. To perform a numerical simulation for explaining the strain-rate effects observed in the experiment, we set up a model of a 2D elastic medium containing a straight fault and assume that the frictional force acting on the fault plane follows a rate- and state-dependent friction law. It is found that the numerical results well explain the experimental observations (1)–(3). This indicates that the friction law is useful for the numerical experiments on the slip nucleation process on faults in laboratories. The numerical simulation results further indicate that the spatial distribution of normal stress controls the rupture propagation during the slip nucleation process and that the macroscopic frictional strength decreases with an increase in the degree of nonuniformity in the normal-stress distribution.

Effect of a large outer rise earthquake on seismic cycles of interplate earthquakes : A model study

A numerical simulation of the sliding process of a plate interface is performed to investigate the effect of stress perturbation due to a large outer rise earthquake on seismic cycles of large interplate earthquakes in a subduction zone. A two-dimensional model of uniform elastic half-space is set up for repeatedly occurring large interplate earthquakes by assuming that the frictional stress on the plate interface obeys a rate- and state-dependent friction law derived from laboratory experiments. An increase in shear stress on a shallow part of the plate interface due to a tensional outer rise earthquake promotes aseismic sliding on the shallower aseismic zone of the plate interface. This aseismic sliding advances the occurrence time of the next interplate earthquake. On the other hand, a compressional outer rise earthquake tends to delay the occurrence time of the next interplate earthquake. The magnitude of time advance or delay of the next interplate earthquake depends on the time of the outer rise earthquake in a seismic cycle due mainly to the complex behavior in propagation of aseismic sliding governed by the rate- and state-dependent friction law.

Microfracture processes in the breakdown zone during dynamic shear rupture inferred from laboratory observation of near-fault high-frequency strong motion

High-frequency velocities are measured during stick-slip motion in the immediate vicinity of a fault in a granite sample to reveal the microscopic process taking place in the breakdown zone defined in the slip-weakening model. It is found that 1) the onset time of the observed strong motion approximately coincides with the local rupture onset time, 2) the observed near-fault high-frequency strong-motion duration is approximately proportional to the local breakdown time, and 3) the power spectra of strong motions exhibit significant amplitudes at frequencies above the value offmax, wherefmax is a cut-off frequency relevant to rupturing the breakdown zone. These observations suggest that the high-frequency motion would be due to the incoherent brittle microfracture whose characteristic scale is much shorter than the breakdown zone size. We present a stochastic fault model to synthesize the near-fault high-frequency velocity waveforms. In the model, a number of small circular subfaults are distributed randomly on the fault and the rupture onset time of an individual subfault is assumed to be random. The main features of the observed velocity waveforms are well explained by this numerical modeling. It is concluded that approximately half of the total energy of high-frequency elastic waves observed at a point is radiated from the propagating breakdown zone. We emphasize the importance of the observation of near-fault high-frequency strong motions for large shallow earthquakes.

Effect of fault bend on the rupture propagation process of stick-slip

Abstract An experimental study of stick-slip is performed to examine the effect of a fault bend on the dynamic rupture propagation process. A granite sample used in the experiment has a pre-cut fault that is artificially bent by an angle of 5.6° at the center of the fault along strike, and accordingly the fault consists of two fault segments. The rupture propagation process during stick-slip instability is investigated by analyzing the records of shear strain and relative displacement measured with strain gauge sensors together with the hypocenters of AE (acoustic emission) events detected with piezoelectric transducers. The observed rupture propagation process of typical stick-slip events is as follows. (1) The dynamic rupture started on a fault segment is stopped near the fault bend. (2) The rupture propagation is restarted near the bend on the other fault segment 10.8 ms to 3.5 s after the stop of the first rupture. The delay time of the second rupture decreases with an increase in the slip amount of the first rupture or a decrease in the normal stress acting on the fault segment where the second rupture started. (3) The restarted rupture is not arrested by the presence of a fault bend, and slip occurs over the entire fault. We theoretically analyze the stress concentration near the fault bend to find that the normal stress produced by the preceding slip near the fault bend plays an important part in controlling the rupture propagation. A numerical simulation based on a rate- and state-dependent friction law is performed to interpret physically the retarded rupture in the experiment. The observed time interval of 10.8 ms to 3.5 s between the first rupture and the second is explained by the numerical simulation, suggesting that the rate- and state-dependence of rock friction is a possible mechanism for the retarded rupture on the fault.

A stress-corrosion model for strain-rate dependence of the frictional strength of rocks

This Technical Note presents a simple model of the time-dependent fracture of asperities on sliding surfaces to explain the experimental results obtained for the strain-rate dependence of friction strength. A model is derived taking into account brittle fracture of asperities in two dimensions with an infinite periodic array of cracks. Kato et al. had examined the strain- rate dependence of maximum frictional strength prior to unstable slip by biaxially compressing a large-scale granite sample with a 40cm long precut fault. The results from this work are presented and compared with the theoretical equation. Predictions that the friction strength is proportional to the 1/(n+1)th power of the strain-rate, where n is the stress corrosion index, are borne out. The experimental values of the power index however are a third to a half that of the theoretical ones. Reasons for this are put forward and areas for future research suggested.

Seismic observations at a seismic gap in the eastern margin of the Japan Sea using ocean bottom seismometers

Abstract The eastern margin of the Japan Sea is a nascent convergent plate boundary. Previous studies proposed the existence of a seismic gap along this boundary between 39°N and 40°N. The trend of this gap is reported by Ohtake (Island Arc 4, 156–165, 1995) to be north-northwest to south-southeast, but by Ishikawa (Gekkan Kaiyo, Suppl. 7, 102–107, 1994) and Matsuzawa (Prog. Abstr., Seismol. Soc. Jpn. 2, B92, 1995) to be north-northeast to south-southwest. During one month ocean bottom seismic observations were conducted using nine ocean bottom seismometers to investigate seismicity in and around the seismic gap area in detail. The observations revealed that the earthquake epicentral distribution had an echelon shape and could be divided into three groups. These groups have a north-northeast to south-southwest trend. This trend is consistent with the fault system in this area, which was formed by the back-arc spreading in the Early to Middle Miocene. This suggests that previously formed tectonic structures affect the present seismo-tectonics and that this area has weak planes with a north-northeast to south-southwest trend.