1. From the trench to the seismogenic zone: Establishing links between low-T metamorphism, fluid pressure, and fault stability
Demian Saffer, Penn State Univ.
SEIZE-Subfac workshop
Heredia, Costa Rica
June, 2007

2. Outline 1. Overview: - Criteria for unstable slip
- Roles of rock properties, fluid pressure
2. Low-T metamorphic processes
- influence on frictional behavior
- influence on fluid pressure
- links to geochemical signals
3. Compaction- and dehydration-driven pore
pressure: spatial distribution and links to
the updip limit

3. SW Nankai Subduction Zone
Parkfield, CA Seismicity
20%
aseismic 5 5
seismogenic
zone 10 10 15
aseismic
The seismogenic zone is defined by the transitions from stable to unstable frictional deformation
Does the proposed process actually occur where suspected?
If so, does the proposed process have the hypothesized effect on sliding stability?
Goal: Test qualitative interpretations and observations that provide only circumstantial evidence for processes that control the onset of unstable slip, through a combination of targeted experiments and modeling.

4. Does the process have the hypothesized effect on sliding stability?
SW Nankai Subduction Zone
Parkfield, CA Seismicity
20%
aseismic 5 5
seismogenic
zone 10 10 15
aseismic
Does the process have the hypothesized effect on sliding stability?
Does the process occur where proposed/suspected?
Does the hypothesis successfully explain observations at other margins?
Does the proposed process actually occur where suspected?
If so, does the proposed process have the hypothesized effect on sliding stability?

5. Simple Spring-Slider System: Force BalanceK F s f x x´
Ffriction = Fnormal * m
Fspring = K * x
after Scholz (2003)

6. Simple Spring-Slider System: Force BalanceK F s f x x´
If frictional resistance decreases more
rapidly than force from spring, forces are not balanced (block can accelerate): B C
Force
Displacement
Slope = -K
Slip
m s x´ x f
dFspring/dx = -K
dFfriction/dx = (dm/dx)*Fnormal
after Scholz (2003)

7. (a-b) < 0 (a-b) > 0 Slip weakening behavior:
measured by velocity-stepping
during shearing experiments
(a)
(a-b) = D µ/ D
ln(V)
SAFOD B1 s n
(a-b) < 0
(MPa) 5
0.01
Coefficient of Friction
100
(a-b) > 0
SAFOD B21 Dc 10 30
100
Load Point
Velocity (µm/s)
Slip
200 µm
p690, p692
Marone (2006)

8. Translated to Rate-and-State Friction Framework:
Change in frictional
resistance with slip
given by: Dc
sn’ (a-b)
Unstable if:
n‘ (a  b) Dc
K <

9. Stability parameter: z = sn’ (a-b)
Stability Criterion
Stability parameter: z = sn’ (a-b)
Rock properties: more negative (a-b) increases
tendency for instability
Pore pressure dissipation: Higher effective stress increases tendency for instability
Thermal pressurization/melting can explain slip weakening, but not nucleation!

10. Hypotheses invoking low-T metamorphism and fluid pressure
After Moore & Saffer (2001)

11. Conceptual model for Costa Rican Margin
Von Huene et al. (2004)

12. Illite powder exhibits only velocity strengthening
Does clay transformation cause a transition to unstable slip? Isolating effect of mineralogy
Smectite powder exhibits both velocity weakening and velocity strengthening
Illite powder exhibits only velocity strengthening

13. Does clay transformation cause a transition to unstable slip
Does clay transformation cause a transition to unstable slip? Isolating effect of mineralogy
Smectite exhibits velocity weakening at low normal stress and velocity strengthening at higher normal stress (for v < 20 micron/s)
Illite exhibits velocity strengthening for all normal stresses and velocities studied
(Saffer, Frye, Marone, and Mair, GRL 2001)
Saffer & Marone (2003)

14. 25% illite in mixed-layer clays 61% illite in mixed-layer clays
What about cementation and fabric that accompany low-T diagenesis & clay transformation?
Triaxial Experiments on intact “wafers” of sediment:
preserve cement, fabric, and porosity
25% illite in mixed-layer clays
estimated T = ~75 C
61% illite in mixed-layer clays
estimated T = ~140 C
McKiernan, Saffer, & Lockner (2005)

15. Implications:
Clay transformation and natural diagenetic processes in bulk sediment do not appear to cause a transition to negative (a-b)…based on experimental data to date.
Therefore, consolidation, mineralization, and evolution of fault zones with high shear strain may be important in the onset of unstable slip (Moore et al., 2007; Ikari et al., 2007).

16. Role of tectonic loading, drainage, and fluid pressure on effective stress
Vertical drainage of subducted section is a robust trend
for several margins
Behavior can be approximated by a simple model of
consolidation
Saffer (2007)
Saffer, 2006

17. Lab data to constrain hydraulic diffusivity
Saffer (2005)

18. Drainage of the subducted section
updip extent of Costa Rica
micro-seismicity 10 20 30 40 50 60
Distance from trench (km)
updip extent of 1946 Nankai
coseismic rupture
Saffer (2007)

19. Correlation with fault behavior: Nankai
Bangs et al. (2004)

20. distance from trench (km) 50 40 30 20 10 1 Cv = 1x10 m s1
Cv = 1x10
-10 m 2 s -1
0.8
Cv = 1x10 -9 m 2 s -1
0.6
Underconsolidation Ratio (U)
0.4
Cv = 3x10 -9 m 2 s -1
Cv = 1x10 -8 m 2 s -1
0.2
décollement down-stepping
Bangs et al. (2004)

21. Effects of metamorphism on fluid pressure
Thermal Model
Compaction
Dehydration
Bound. Conds
Fluid Sources
Fluid Pressure
Permeabilities

22. Illitization model calibration: Nankai ODP Sites
Huang (1993) and Pytte & Reynolds (1988) kinetic expressions
808
1174
1173
-1200
-1000
-800
-600
-400
-200 20 40 60 80
100
-1000
-800
-600
-400
-200 20 40 60 80
100
-700
-600
-500
-400
-300
-200
-100 20 40 60 80
100
100
200
200
200
400
400
300
Depth (mbsf)
600
600
400
800
800
500
1000
600
1000
1200
700 20 40 60 80 20 40 60 80 20 40 60 80
% Illite in I/S mixed Layer Clays

23. Application to Nankai Margin
x 10-14 4
updip extent of 1946
coseismic rupture 3
compaction-driven sources
Fluid Production
(VH2O / Vsediment s-1) 2
Muroto
Ashizuri 1 10 20 30 40 50
distance landward from trench (km)

24. Application to Nicoya Margin

25. Integrated 2-D Model for Nicoya Margin
0.2
0.4
0.6
0.8
1.0 20 40 60 80
distance from trench (km) 10 30
depth (km)
l* along decollement
Modeled Pore Pressure
Interplate Earthquakes
after Spinelli et al., 2004

26. Modeled heating and metamorphism:
Also provides insight
into source regions for
geochemical signals
This is a key constraint
for hydrologic models
used to estimate pore
pressure at depth
100 10
[B] (mMol) 1
0.1 40 30
d11B (‰) 20 10

27. Implications: Simulated dissipation of fluid sources and
pressure correlate broadly with updip limit
Negative velocity-dependence is required
for unstable slip, but causes remain unknown
Consolidation and metamorphism intertwined
as potential causes for both transition to
negative (a-b) and increasing sn’.

28. Where to go from here? Causes for transition to velocity-
weakening behavior: experiments on
appropriate natural samples
Permeability measurements for forward
models of pore pressure
Marriage of hydrologic models with
geochemical data to better constrain
hydraulic architecture, pressures at
depth

29. Key Unknowns: Role of seamounts? (hydraulic architecture,
distribution of incoming sediment, bsmt
alteration)
Role of fluids at greater depths (>> 150 C)?

30. Heterogeneity in the Seismogenic Zone
ODP Drilling in Upper Aseismic Zone, with future Goals in the Seismogenic Zone
Areas of concentrated
slip or “asperities”
Seismogenic Zone: Were slip occurs during the earthquake. Note that this slip may be patchy or concentrated in “asperities”. Deal with the mechanical definition of the seismogenic zone later.
after Bilek et al., 2004; Lay and Bilek (2007)

31. Drilling will sample the transition
Stable Slide Region
Stick-slip Region
???
after Saffer & Marone (2003); Tobin (2007)