improved understanding of secular and transient deformation in Southern California and loading of How can the CRM contribute to seismogenic faults?

Transcription

1 How can the CRM contribute to improved understanding of secular and transient deformation in Southern California and loading of seismogenic faults? Yuri Fialko Institute of Geophysics and Planetary Physics Scripps Institution of Oceanography University of California San Diego

2 CRM in SCEC5 σ ε = F ( σ ) ε

3 Secular velocities (CMM4)

4

5 Hector Mine Transient deformation (e.g., post-seismic relaxation) Freed et al., 2012 El Mayor Gonzalez et al., 2014

6 Temporal evolution of surface deformation - Rapid initial transient followed by a more gradual decay -Difficult to fit assuming exponential dependence (not consistent with linear Maxwell viscoelastic behavior) -Possible explanations: Bi- (or multi-) viscous rheology (Pollitz, 2005) Power-law rheology (Freed and Burgmann, 2004) Rate-and-state friction (or some other form of localized creep, e.g., Hearn et al., 2002; Perfettini and Avouac, 2005) Post-Landers CGPS data (Fialko, 2004)

7 Fault perpendicular distance (km) Surface velocity due to a strike-slip fault 50 LOS velocity (mm/yr) CCF 100 (a) Homog., vertical SAF, no CF resid. = Fault perpendicular distance (km) SAF EF Southern San Andreas Fault LOS velocity (mm/yr) CCF EF 100 Homo 50 0 Fault perpendicular distance Lindsey & Fialko CF et al., 2007] and geometry of locked faults in our preferred model (black lines). [Suess and Shaw, 2003; Plesch et al., 2009], with relocated seismicity (black d Figure 2. Shear modulus computed from the SCEC regional velocity model CV Depth (km) 5 SAF

8 vs Savage and Burford, 1973 Turcotte and Spence, 1974 Fault-block vs Viscous sheet Crème Brûlée vs Jelly Sandwich Bottom-driven vs Side-driven Geodetic slip rates and fault locking depths depend on model assumptions Difference between the models results from oversimplified model assumptions? Elsasser, 1969 Savage and Prescott, 1978 Li and Rice, 1987 Johnson and Segall, 2004

9 Earthquake cycles on a rate-and-state fault Kaneko and Fialko, 2011 Tse and Rice, 1987 Lapusta et al., 2001

10 surface velocity slip rate

11 2 cm/yr Upper Mantle Lower Crust Upper Crust 300 km Fault (0-17 km) 75 km y x z 12 km 18 km 35 km 100 km periodic Visco-elastic EQ cycle model free pinned

12 Surface velocities mo 2 yr 10 yr 50 yr 180 yr a) M c) DNK e) DCK g) DCS b) M d) WNK 0 Distance from fault (km) f) WCK h) WCS 0 linear visco-elastic power-law Fault parallel velocity (cm/yr)

13 ! ε T = C σ n exp! εσ ' ( ) Realistic rheologies lead to localization Q & $ RT % viscosity temperature velocity Yuen et al., 1978; Fleitout and Frodivaux, 1980; Turcotte and Schubert, 2002

14 Strain localization mechanisms! Thermo-mechanical coupling! Grain size reduction! Foliation / fabric development! Mineral alteration! Pressure solution! All lead to the development of shear zones and strength reduction

15 Grain size evolution!ε v =!ε D +!ε G = A D σ n D + A G d m σ n G dislocation creep diffusion creep n D ~ 3 n G ~ 1 m ~ 3! d+ = B p 1 d 1 p " exp H G $ # RT ij % ' & static grain growth! d = λd!ε G dynamic recrystallization de Bresser et al. (1998; 2001) Hall and Parmentier (2003)

16 Effects of dynamic recrystallization Dynamic recrystallization TMC Dynamic recrystallization + TMC strain rate

17 Strain localization at 20 km depth Velocity, mm/yr 5 0 GS only TC only GS+TC Distance from fault, km Figure 5: Profiles of fault-parallel velocity at depth of 20 km, for the deformation mechanis illustrated in Figure 4: grain size evolution (red line), thermomechanical coupling (blue lin

18 Foliation Strain-induced separation of weak and strong mineral phases; development of anisotropic fabric N!ε =!ε φ tot i i i=1 N φ =1 i i=1

19 Aggregate vs individual phase rheologies Anorthite, wet Diopside, wet Two phase mixture, wet Effective viscosity, Pa s Stress, Pa Figure 2: E ective viscosities corresponding to the dislocation creep for pure wet anorthite (blue), pure wet flow diopside law parameters (red), and a wet from 50% Dimanov anorthite-50% and Dresen diopside mixture (1995) (green) over a range of shear stresses at a temperature of 500 C. Flow law parameters were taken from Dimanov and Dresen [2005].

20 Deformation (and strength) is controlled by the weakest phase Velocity, m/s Anorthite, wet Diopside, wet Two phase mixture, wet Distance, cm Figure 3: Distribution of velocity across a 30 cm-thick layer of simulated material composed of three laminated flow law rheologies: parameters 0-10 from cm: 5 cm Dimanov pure wet and anorthite Dresen (blue) (1995) and 5 cm pure wet diopside (red); cm: 2.5 cm pure wet anorthite (red), 2.5 cm pure wet diopside (blue), 5 cm wet 50% anorthite-50% diopside mixture (green); 10 cm wet 50% anorthite-50% diopside mixture (green), subjected to a shear stress of 10 MPa at a temperature of 500 C. All phases deform by dislocation creep, with flow law parameters from Dimanov and Dresen [2005].

21 Mylonite zone (Ontario, Canada) Thurber et al. (2006) Shelly (2010) Zhu (2000)

22 Gueydan et al., 2014

23 Field geology & laboratory view of a major crustal fault Scholz, 1988

24 simple conceptual models of interseismic deformation (both wrong)

25 vs

26 Effect of the model size

27 vs?? bottom tractions?

28 "!" )" (" '" #" $" %" 9/:/"/78;/4<* 9/:/'$/78;/4<* 9/:/"/78;/=<* 9/:/'$/78;/=<* Stresses predicted by EQ cycle models employing lab rheology Wet vs dry makes a big difference!!" #!" $!" %!" & *+,-../ /0783

29 Stress: constraints from the rock record Behr and Platt, 2011 G. Hirth

30 DCS WCS Stress in the lithosphere as a function of plate velocity: faster = weaker σ ε -1! Slip rate (cm/yr) Stress (MPa)

31 Challenges for CRM! Effective rheology: linear or non-linear?! Are steady-state constitutive laws appropriate ( transient rheology?)! Strain localization: how to quantify and incorporate in the model?! Need better understanding of a transition from rate-state friction to ductile deformation! Water content: as important as temperature. CWM?