"Subduction-zone seismicity and emerging new problems in fault mechanics"
| Cosa | Seminari seminari |
|---|---|
| Quando | 22/11/2010 da 14:30 al 23:55 |
| Dove | sala conferenze INGV Roma |
| Aggiungi l'evento al calendario |  vCal (Windows, Linux) iCal (Mac OS X) |
22 NOVEMBRE 2010 ore 14.30 | Prof. T. Shimamoto | Sala Conferenze Roma | Sede Centrale
A seminar at INGV, Rome (22 Nov. 2010)
Subduction-zone seismicity and emerging new problems in fault mechanics
Toshihiko Shimamoto
State Key Laboratory of Earthquake Dynamics, Institute of Geology
China Earthquake Administration
P. O. Box 9803, Beijing 100029, China
E-mail:Questo indirizzo email è protetto dagli spambots. È necessario abilitare JavaScript per vederlo. , Questo indirizzo email è protetto dagli spambots. È necessario abilitare JavaScript per vederlo.
1. Introduction
A traditional view of velocity weakening as a required property for seismogenic fault motion was challenged for the 1999 Chi-Chi earthquake based on laboratory data indicating velocity strengthening for the southern part of Chelungpu fault which displaced even more than the northern part (Tanikawa and Shimamoto (2008, JGR). They proposed that rate-and-state friction at slow slip rates controls the earthquake nucleation, whereas intermediate to high-velocity friction dictates the growth processes into a large earthquake. Noda & Lapusta (2009, JpGU meeting) demonstrated by dynamic modeling that such a scenario is indeed possible. This changed my view on subduction-zone seismicity completely and I proposed that the drilling project has to be reorganized with a renewed view (Shimamoto, 2009, Seismol. Soc. Japan Meeting, Kyoto; AGU, 2009, San Francisco). This was extended to deeper portion of subduction zone, slow-slip regime and transitional regime between megathrust and slow slip regimes (Shimamoto, JpGU 2010 meeting). I would like to elaborate in my talk the seven major problems in fault mechanics for better understanding of subduction-zone seismicity.
2. Seven tasks in fault mechanics
The first 3 tasks and next 4 tasks are related to shallow and deep parts of subduction zones, respectively. We should challenge those tasks in the next few years.
Task 1: High-velocity (HV) friction of faults to evaluate their response to earthquake rupture coming from depths. A megathrust earthquake nucleates at a deeper part for well-coupled plate interface such as Nankai Trough, so HV property needs to be determined to analyze the response of shallow faults or plate interface. In particular, initial peak friction needs to be studied systematically under dry and wet conditions.
Task 2: Friction and fracture experiments to determine velocity dependence of faults and post-failure curve. Deformation of accretionary prism itself must be causing ultra-low-frequency earthquakes in shallow parts. Also, fault zones are still evolving at shallow depths and fracturing experiments are equally important as are friction experiments.
Task 3: Reexamination of exhumed accretionary prism such as Shimanto belt with a renewed view that background deformation of accretionary prism itself is overprinted by impulse-like deformation due to rupture propagation from depths. The most important task is to identify major deformation mechanisms in megathrust faults that cannot be reached by drilling even with Chikyu.
Task 4: High-temperature and ultralow effective pressure (Pe) friction experiments to understand slow slip and nonvolcanic slow-frequency tremors in the transitional regime. Recent modeling of slow slip by several groups strongly suggests that Pe is on the order of several MPa in the slow-slip regime. Some fault rocks may retain brittle frictional properties even at high temperatures, but other rocks may not. No data at present under such conditions and are urgently needed!
Task 5: Deformation mechanisms along megathrust faults, particularly evaluation of the significance of pressure solution. Pressure solution is likely to be important to produce flow deformation in metamorphic schists. But this process exhibit velocity strengthening when fully operational and cannot lead to earthquake nucleation. But megathrust earthquake still can occur if there is a nucleation site, say at the base of megathrust. Task 3 will provide a clue to this problem.
Task 6: Studies on hydrofractures, permeability and fracture seal in metamorphic environments which are needed to analyze pore-pressure (Pp) evolution in subduction zones. How unusually high Pp can be maintained in slow slip regime is a difficult but an exciting problem. Pp may be the most important factor in delineating the megathrust and slow-slip regimes.
Task 7: Dynamic analysis of slow slip and megathrust earthquake cycles using realistic fault properties are needed to understand how a megathrust earthquake initiates. Recent modeling of Matsuzawa et al. (2009, JpGU and AGU) brought about changes in frequency of slow slip prior to a megathrust earthquake. N. Lapusta and H. Noda have developed a fully dynamic software incorporating realistic fault properties and is ideal for such analyses.
As for task 4 (controlling frictional properties in the slow slip regime), there was some progress on an empirical constitutive law linking fault constitutive law to fully plastic flow law (see abstract submitted to AGU). This simple law fits experimental data very well and has predictive power on friction-flow properties in the intermediate regime, relevant to the slow slip regime. The law will be a guide for planning experiments in the future. I and Noda are now working on fine details of this and I plan to include this in this talk.
Abstract submitted to AGU, San Francisco, Dec. 2010
A new brittle to plastic constitutive law and its implications for subduction-zone seismicity
Toshihiko Shimamoto and Hiroyuki Noda
Institute of Geology, China Earthquake Administration
Seismological Laboratory, California Institute of Technology
Establishment of a constitutive law from brittle deformation to plastic flow has long been a task for solving problems such as modeling earthquake generation and plate interactions. Although solution-precipitation creep may be involved with deformation in fluid-rich environments, we focus here on how to link a friction law in the brittle regime to flow law in fully plastic (or pressure-independent) regime. One way to connect friction and flow laws is to use a linear combination of the two, but this cannot account for the behavior in the intermediate regime (Noda and Shimamoto, 2010, GRL). Here we explore a simple empirical equation, tss/tflow = tanh (mse/tflow), where tss and se are the shear resistance and the effective normal stress on a fault or on a shear zone, respectively, tflow is the flow stress as calculated from a flow law under given conditions, and m is the frictional coefficient. Note that this equation reduces to friction law when the flow stress is much greater than the frictional resistance and to flow law when friction is much larger than the flow stress. It holds for steady state if steady-state friction and steady-state flow stress are used. It also describes transient and steady-state behavior if a rate-and-state constitutive law and flow law including transient behavior are used. A rate and state flow law was proposed by Noda and Shimamoto (2010, GRL) and this can be used for the latter.
So far halite is the only material for which a complete transition from frictional slip to fully-plastic flow was achieved in experiments. The above equation was proposed originally by Shimamoto (2004) who showed that the equation describes the observed behavior from frictional slip to flow with increasing normal stress at different temperatures. We show here that it agrees well with velocity dependence of shear resistance with slip rate or shear strain rate with increases in normal stress and temperature (Kawamoto and Shimamoto, 1997). We are now exploring transient behaviors from brittle to plastic regimes.
Our law predicts a change in velocity weakening to velocity strengthening with increasing normal stress and temperature, in agreement with experimental data. Stick-slip indeed disappears at this point in experiments. Thus the lower bound of seismogenic zone can be regarded as the velocity weakening to velocity strengthening transition as often treated in earthquake modeling. However, this transition is caused by partial involvement of plastic flow, not simply by a change in velocity dependence of friction law. Thus modeling of generation of large and great earthquakes has to be done with a friction to flow law. Another important implication is for slow slip in subduction zones. It is highly likely that fluid pressure is nearly equal to the lithostatic pressure in slow slip regimes. A simple strength profile indicates that the brittle regime expands with increasing pore pressure. Our law can predict the deepening of the transition depth from velocity weakening to strengthening, which probably defines the lower bound of slow slip regime, although the width of plastic shear zone at depths remains as an uncertain parameter.
Subduction-zone seismicity and emerging new problems in fault mechanics
Toshihiko Shimamoto
State Key Laboratory of Earthquake Dynamics, Institute of Geology
China Earthquake Administration
P. O. Box 9803, Beijing 100029, China
E-mail:
1. Introduction
A traditional view of velocity weakening as a required property for seismogenic fault motion was challenged for the 1999 Chi-Chi earthquake based on laboratory data indicating velocity strengthening for the southern part of Chelungpu fault which displaced even more than the northern part (Tanikawa and Shimamoto (2008, JGR). They proposed that rate-and-state friction at slow slip rates controls the earthquake nucleation, whereas intermediate to high-velocity friction dictates the growth processes into a large earthquake. Noda & Lapusta (2009, JpGU meeting) demonstrated by dynamic modeling that such a scenario is indeed possible. This changed my view on subduction-zone seismicity completely and I proposed that the drilling project has to be reorganized with a renewed view (Shimamoto, 2009, Seismol. Soc. Japan Meeting, Kyoto; AGU, 2009, San Francisco). This was extended to deeper portion of subduction zone, slow-slip regime and transitional regime between megathrust and slow slip regimes (Shimamoto, JpGU 2010 meeting). I would like to elaborate in my talk the seven major problems in fault mechanics for better understanding of subduction-zone seismicity.
2. Seven tasks in fault mechanics
The first 3 tasks and next 4 tasks are related to shallow and deep parts of subduction zones, respectively. We should challenge those tasks in the next few years.
Task 1: High-velocity (HV) friction of faults to evaluate their response to earthquake rupture coming from depths. A megathrust earthquake nucleates at a deeper part for well-coupled plate interface such as Nankai Trough, so HV property needs to be determined to analyze the response of shallow faults or plate interface. In particular, initial peak friction needs to be studied systematically under dry and wet conditions.
Task 2: Friction and fracture experiments to determine velocity dependence of faults and post-failure curve. Deformation of accretionary prism itself must be causing ultra-low-frequency earthquakes in shallow parts. Also, fault zones are still evolving at shallow depths and fracturing experiments are equally important as are friction experiments.
Task 3: Reexamination of exhumed accretionary prism such as Shimanto belt with a renewed view that background deformation of accretionary prism itself is overprinted by impulse-like deformation due to rupture propagation from depths. The most important task is to identify major deformation mechanisms in megathrust faults that cannot be reached by drilling even with Chikyu.
Task 4: High-temperature and ultralow effective pressure (Pe) friction experiments to understand slow slip and nonvolcanic slow-frequency tremors in the transitional regime. Recent modeling of slow slip by several groups strongly suggests that Pe is on the order of several MPa in the slow-slip regime. Some fault rocks may retain brittle frictional properties even at high temperatures, but other rocks may not. No data at present under such conditions and are urgently needed!
Task 5: Deformation mechanisms along megathrust faults, particularly evaluation of the significance of pressure solution. Pressure solution is likely to be important to produce flow deformation in metamorphic schists. But this process exhibit velocity strengthening when fully operational and cannot lead to earthquake nucleation. But megathrust earthquake still can occur if there is a nucleation site, say at the base of megathrust. Task 3 will provide a clue to this problem.
Task 6: Studies on hydrofractures, permeability and fracture seal in metamorphic environments which are needed to analyze pore-pressure (Pp) evolution in subduction zones. How unusually high Pp can be maintained in slow slip regime is a difficult but an exciting problem. Pp may be the most important factor in delineating the megathrust and slow-slip regimes.
Task 7: Dynamic analysis of slow slip and megathrust earthquake cycles using realistic fault properties are needed to understand how a megathrust earthquake initiates. Recent modeling of Matsuzawa et al. (2009, JpGU and AGU) brought about changes in frequency of slow slip prior to a megathrust earthquake. N. Lapusta and H. Noda have developed a fully dynamic software incorporating realistic fault properties and is ideal for such analyses.
As for task 4 (controlling frictional properties in the slow slip regime), there was some progress on an empirical constitutive law linking fault constitutive law to fully plastic flow law (see abstract submitted to AGU). This simple law fits experimental data very well and has predictive power on friction-flow properties in the intermediate regime, relevant to the slow slip regime. The law will be a guide for planning experiments in the future. I and Noda are now working on fine details of this and I plan to include this in this talk.
Abstract submitted to AGU, San Francisco, Dec. 2010
A new brittle to plastic constitutive law and its implications for subduction-zone seismicity
Toshihiko Shimamoto and Hiroyuki Noda
Institute of Geology, China Earthquake Administration
Seismological Laboratory, California Institute of Technology
Establishment of a constitutive law from brittle deformation to plastic flow has long been a task for solving problems such as modeling earthquake generation and plate interactions. Although solution-precipitation creep may be involved with deformation in fluid-rich environments, we focus here on how to link a friction law in the brittle regime to flow law in fully plastic (or pressure-independent) regime. One way to connect friction and flow laws is to use a linear combination of the two, but this cannot account for the behavior in the intermediate regime (Noda and Shimamoto, 2010, GRL). Here we explore a simple empirical equation, tss/tflow = tanh (mse/tflow), where tss and se are the shear resistance and the effective normal stress on a fault or on a shear zone, respectively, tflow is the flow stress as calculated from a flow law under given conditions, and m is the frictional coefficient. Note that this equation reduces to friction law when the flow stress is much greater than the frictional resistance and to flow law when friction is much larger than the flow stress. It holds for steady state if steady-state friction and steady-state flow stress are used. It also describes transient and steady-state behavior if a rate-and-state constitutive law and flow law including transient behavior are used. A rate and state flow law was proposed by Noda and Shimamoto (2010, GRL) and this can be used for the latter.
So far halite is the only material for which a complete transition from frictional slip to fully-plastic flow was achieved in experiments. The above equation was proposed originally by Shimamoto (2004) who showed that the equation describes the observed behavior from frictional slip to flow with increasing normal stress at different temperatures. We show here that it agrees well with velocity dependence of shear resistance with slip rate or shear strain rate with increases in normal stress and temperature (Kawamoto and Shimamoto, 1997). We are now exploring transient behaviors from brittle to plastic regimes.
Our law predicts a change in velocity weakening to velocity strengthening with increasing normal stress and temperature, in agreement with experimental data. Stick-slip indeed disappears at this point in experiments. Thus the lower bound of seismogenic zone can be regarded as the velocity weakening to velocity strengthening transition as often treated in earthquake modeling. However, this transition is caused by partial involvement of plastic flow, not simply by a change in velocity dependence of friction law. Thus modeling of generation of large and great earthquakes has to be done with a friction to flow law. Another important implication is for slow slip in subduction zones. It is highly likely that fluid pressure is nearly equal to the lithostatic pressure in slow slip regimes. A simple strength profile indicates that the brittle regime expands with increasing pore pressure. Our law can predict the deepening of the transition depth from velocity weakening to strengthening, which probably defines the lower bound of slow slip regime, although the width of plastic shear zone at depths remains as an uncertain parameter.