Fluids, subduction decollement, cold seeps Pierre Henry CNRS, Collge de France

1
Fluids, subduction decollement, cold seepsPierre
Henry (CNRS), Collège de France CEREGE,
Aix-en-Provence

• Introduction
• Fluids in the subduction factory
• Heat flow anomalies in Barbados Trough
• Physical concepts
• Subduction from top to as deep you can get
• Aseismic decollement
• Case studies Barbados and Nankai
• Solitary wave
• Silent slip
• Seismogenic zone
• Drilling projects
• Hydrogeologic aspects
• Sum up
• Cold seeps on active faults
• Nankai dilema
• Preliminary results

2
(No Transcript)
3
Barbados accretionary wedge Heat flow anomalies
in the trench
Etopo5 bathymetry
Background heat flow in Barbados Trough is 50-60
mW/m2 (100 Ma oceanic lithosphere)
4
High heat flow anomalies and basement highs
Gravity data, Sandwell version 7.2
Fluid migration toward basement highs ?
5
Fluid and fault activity

• Fluid pressure and friction laws
• Normal stress reduction
• Cause of fault weakness ?
• Cause of seismic/aseismic transition ?
• Pressure transients during sliding
• Dilatancy/compaction
• Water thermal expansion
• Fault zone permeability
• Core Barrier
• Damage zone Channel
• Episodic events
• Inactive faults are seals
• Sibson (1981) distinguishes  seismic pumping 
  and  valve  based of the causal relationship
  with earthquakes

(Caine et al., 1996)
6
References (1)

• Byerlee, J. D. (1990). "Friction, overpressure
  and fault normal compression." GRL 17 2109-2112.
• Rice, J. (1992). Fault stress states, pore
  pressure distributions, and the weakness of the
  San Andreas Fault. Fault mechanics and transport
  properties of rocks. B. Evans and T.-F. Wong.
  London, Academic 475-503.
• Sleep, N. H. and M. L. Blanpied (1994). "Ductile
  creep and compaction a mechanism for transiently
  increasing fluid pressure in mostly sealed fault
  zones." Pure and Applied Geophysics 143(1-3)
  9-40.
• Scholz, C. H. (1998). "Earthquakes and frictions
  laws." Nature 391 37-42.
• Segall, P. and J. R. Rice (1995). "Dilatancy,
  compaction, and slip instability of a
  fluid-infiltrated fault." JGR 100 22155-22171.

• Sleep, N. H. (1995). "Frictional heating and the
  stability of rate and state dependent frictional
  sliding." GRL 22 2785-2788.
• Caine, J.S., Evans, J.P., and Forster, C.B.,
  1996, Fault zone architecture and permeability
  structure Geology, v. 24, p. 1025-1028.
• Barton, C.A., Zoback, M.D., and Moos, D., 1995,
  Fluid flow along potentially active faults in
  crystalline rock Geology, v. 23, p. 683-686.
• Carson, B., and Screaton, E.J., 1998, Fluid flow
  in accretionry prisms evidence for focused,
  time-variable discharge Reviews of Geophysics,
  v. 36, p. 329-352.

7
Episodic eventsSibson (1981)  seismic
pumping  and  valve 

• Reversible processes
• Poroelasticity and pressure diffusion
• Response recorded in boreholes away from fault
  zones
• Irreversible processes
• Ground shaking (fluidization and damage)
• Expulsion of water and gas
• Fracturing/sealing cycles
• Minéral veins
• Overpressured fault weakness
• Sliding instability
• Non-linearity of Darcys law k(Pfluid)
• Solitary waves
• aseismic events (ultra-slow)
• Migration of microseismicity
• Migration of hydrocarbons
• Mud volcanoes and phreatomagmatic eruptions

8
References (2)

• Sibson, R.H., 1981, Fluid flow accompanying
  faulting Field evidence and models, in Simpson,
  D.W., and Richards, P.G., eds., Earthquake
  prediction An international review, Volume 4
  American Geophysical Union Maurice Ewing Series,
  p. 593-603.
• Muir-Wood, R., and King, G.C.P., 1993,
  Hydrological signatures of earthquake strain J.
  Geophys. Res., v. 98, p. 22035-22068.
• Miller, S.A., 1996, Fluid-mediated influence of
  adjacent thrusting on the seismic cycle at
  Parkfield Nature, v. 382, p. 799-802.
• Gratier, J.P., Favreau, P., and Renard, F., 2002,
  Modeling fluid transfer along californian faults
  when integrating pressure solution crack sealing
  and compaction process J. Geophys. Res., 108,
  Art. 2140, 2003.
• Husen, S., and Kissling, E., 2001, Postseismic
  fluid low after the large subduction earthquake
  of antofagasta, Chile Geology, v. 29, p.
  847-850.

• Rice, J., 1992, Fault stress states, pore
  pressure distributions, and the weakness of the
  San Andreas Fault, in Fault mechanics and
  transport properties of rocks, edited by B. Evans
  et T.-F. Wong, pp. 475-503, Academic, London.
• Henry, P., 2000, Fluid flow at the toe of the
  Barbados accretionary wedge constrained by
  thermal, chemical, and hydrogeologic observations
  and models J. Geophys. Res., v. 105, p.
  25855-25872.
• Revil, A., 2002, Genesis of mud volcanoes in
  sedimentary basins. A solitary wave-based
  mechanism, Geophysical Research Letters, 29,
  doi10.1029/2001GL014465.
• Miller, S.A., Collettini, C., Lauro, C., Cocco,
  M., Barchi, M. and Kaus, B. J. P., 2004,
  Afterchocks driven by a high pressure CO2 source
  at depth, Nature, 427, 524-427.

9
Hot subduction limits of the seismogenic zone
Hyndman et al. (1997)
10
Decollements evolution of ideas

• Paradigms
• Critical coulomb wedge theory suggested
  quasi-lithostatic pore pressures in accretionary
  wedges and decollement (Dahlen, 1984)
• Heat flow and chemical anomalies in the trench
  attributed to fluid flow along the decollement
  (Langseth et al., 1990 Le Pichon et al., 1990
  Kastner et al., 1991 Moore and Vrolijk, 1992)
• Decollement initiation by hydraulic fracturing in
  the trench (Westbrook et Smith, 1983)
• ODP results (Barbados, Nankai)
• Moderate overpressuring (Davis et al., 1997
  Foucher et al., 1997 Screaton et al., 2002
  Saffer, 2003) and low friction coefficient (Kopf
  and Brown, 2003)
• Some heat flow and chemical anomalies have other
  explanations (Henry, 2000 Henry et Bourlange,
  2004)
• Microstructural observations (Labaume et al.,
  1997 Ujiie et al., 2003) and physical
  measurements (Bourlange el al., 2003) suggest
  fluid pressure cycling
• Mecanical decoupling initiates in the trench
  (Owens, 1993 Morgan and Karig, 1993,1995
  Housen, 1997, Henry et al., 2003)

11
Barbados CORK experiment(ODP Leg 156)
lb0.79
lb0.70

• Moderate overpressuring
• Decollement is probably not the source of the
  high heat flow anomaly

12
Transient fluid migration along Barbados
decollement
13
Solitarypressure wave
14
Japanesetectonic context

• Plates
• Kinematics
• Subductions
• Nankai

15
Nankai Leg ODP 190 et 196Rôle des fluides dans
linitiation dun décollement

• CEC, chloride and fluid migration
• In situ smectite-illite reaction (Henry et
  Bourlange, 2004)
• Anisotropy of electrical conductivity and
  deformation
• 10-15 ductile shortening at deformation front
  (Henry et al., 2003)
• Decoupling along the decollement initiates in the
  trench
• Fluid pressure increase
• Compaction and dilatancy in the decollement zone
  (Bourlange et al., 2003)
• Measurements on samples -gt porosity loss
• Measurements in borehole (LWD) -gt fracture
  dilatancy
• Hypothesis
• Coexistence of dilatant et compactive deformation
  modes
• Fluid pressure cycling in the decollement

gt!
16
Nankailow-chloride anomalyLateral fluid flow
along decollement or in situ diagenesis ?
(From Leg 190 Initial reports and Spivak et al.,
2001)
17
smectite -gt illite reaction fluid budget based
on Cation Exchange Capacity(Henry and Bourlange,
EPSL, in press)
18
Nankai Evidence for early decouplingSuggests
extensional state of stress below decollement
10-15 ductile shortening (Henry et al., 2003)
vertical compaction
(see also, on Nankai Owens, 1993 Morgan and
Karig, 1993,1995 on Barbados Housen, 1997)
19
Archie Graph
20
Definitions

• The following parameters are defined from
  electrical conductivity measurements in 3
  orthogonal directions (Sx, Sy and Sz)
• Normalised horizontal conductivity 2 Sx/(SxSy)
• Horizontal anisotropy 200 (Sx-Sy)/(SxSy)
• Vertical anisotropy 200 (Sx/2Sy/2-Sz)/(Sx/2Sy/
  2Sz)
• Anisotropy ratio 100 (Sx-Sy)/(Sx/2Sy/2-Sz)

21
Roses Site 1174
22
Nankai quantifying ductile deformation
March theory applied to the anisotropy of
electrical conductivity 10-15 horizontal
shortening above the decollement associated with
porosity loss
23
compactive and dilatant fractures - RAB image
from Site 808(McNeill, Ienaga et al., 2003)
Frontal thrust
Fault zone
Décollement zone
24
LWD Site 808 (Bourlange et al.,2003)

• Sample porosity
• Porosity computed from porosity log (RAB tool
  model from Revil et al., 1998)
• Fracture conductivity model (m1) applied to the
  decollement

25
Pressure wave equations
Non linear diffusion
Effective pressure
Specific Storage
Permeability function of effective pressure
26
Pressure wave velocity(Rice, 1992)
wedge taper (3-10)
bulk density (1700-2700 kg/m3)
Pressure gradient 0.5-2 Mpa/km
fluid density (1000 kg/m3)
permeability of dilated fault zone (10-14 -10-11
m2)
permeability of closed fault zone (10-19 -10-16
m2)
Wave velocity 10-10000 m/an
Fault zone dilatancy or fracture porosity (4-8)
Viscosity (10-3 Pa.s)

• In 2D, pression diffusion into the formation
  slows the wave. Propagation stops for a
  permeability contrast of 1000 between
  décollement and formation

27
k (m2), t 0
P0
?P/?n0
?P/?n0
?P/?n0
Pinjection 11 Mpa
28
Pf (Pa), t 0
P0
?P/?n0
?P/?n0
?P/?n0
Pinjection 11 gt 12 Mpa à t 0
29
Pf (Pa), t 159 a
30
Pf (Pa), t 476 a
31
Pf (Pa), t 1111 a
32
Pf (Pa), t 2063 a
33
Pf (Pa), t 3333 a
34
Pf (Pa), t 4920 a
35
Pf (Pa), t 6508 a
36
Pf (Pa), t 8095 a
37
Pf (Pa), t 9683 a
38
k (m2), t 9683 a
P0
?P/?n0
?P/?n0
?P/?n0
Pinjection 12 MPa
39
Pressure wave velocity (2D)
40
Triggering and maintaining a pressure wave

• Hypotheses
• Source of decollement fluid is the subducted
  sequence
• Background stress below the decollement is
  extensional
• Fluid flux into the decollement is controlled by
  stress state

41
Silent slip
Dragert et al., 2001
Propagation 5-15 km/day Recurrence
interval 14 months (Miller et al., 2002)
42
Silent slip and fluid

• Propagates too fast to be a solitary wave
• Fluid transfer into or out of the fault zone may
  be one controling process
• Pressure diffusion in connected pore network
  should be high (k 10-18 m2, 1/ja 100 GPa ,
  D3000 m2/an)
• Fluid transfer during one cycle should scale with
  stress variations and drained compressibility
• Dj 1/Kd Ds 100 kPa / 100 GPa 1e-6
• Dj V/t Dt 1/1 Myr 1 yr 1 e-6

43
Seismogenic zone

• Hypotheses
• Systematic, progressive material and state
  changes control the onset of seismogenic behavior
  on subduction thrusts (Hyndmans hypothesis).
• Subduction zone megathrusts are weak faults
  (because of high fluid pressure...)
• Within the seismogenic zone, relative plate
  motion is primarily accommodated by coseismic
  frictional slip in a concentrated zone.
• Physical properties, chemistry, and state of the
  fault zone change systematically with time
  throughout the earthquake cycle (includes pore
  pressure and permeability).
• Splay faults slip during dicrete events
  (tsunamigenic slip) and are hydrogeologically
  connected to the seismogenic zone.
• Projects Nankai (Nantroseize) et Costa Rica
  (CRISP)

44
Nankai seismogenic zone defined by slip
distribution

• GPS
• uniformly locked
• Tsunamis
• heterogeneous sliding

45
Costa Rica seismogenic zone defined from
aftershock distribution
46
Nankai and Costa Rica comparing velocity
structures and seismogenic zones
(Nakanishi et al., 2002)
47
Postseismic fluid flow after the large subduction
earthquake of Antofagasta, Chile (Husen and
Kissling, 2001)
A Early subset shortly after main shock (August
18-37, 1995)
C Proposed explanation
B Late subset a month after main
shock (September 16-27, 1995)
C Entire data set
48
...conceptual model...
49
Summary

• Prisme daccrétion - results
• Fluid sources compaction and diagenesis
• Episodic fluid migration into and along fault
  zones
• Probable relationship between fluid migration,
  aseismic slip and decollement initiation
• Sismogenic zone - testable hypothesis
• Importance of dissolution-recrystalisation
  phenomena
• Self-sealing fault zone
• Important fluid pressure variations
• Ductile zone - Is slow slip controled by fluid ?
• Evaluate the fluid budget of one slip event...

50
NantroseizeNankai subductionTo-Nankairupture
zone
To-Nankai 1944
SFJ
Nantroseize
51
Nantroseize CDP

• Chikyu riser drillship available in 2007
• Phase 1 Reference sites
• Nature and state of entering sediments and
  oceanic crust
• Heat flow and fluid migration in the sediment and
  crust
• Aseismic decollement
• Phase 2 Splay fault
• Properties of the same fault zone at increasing
  P,T conditions
• History (tectonic, seismic, tsunamigenic)
• Hydrogeologic connexion with the seismogenic zone
• Phase 3 Plate boundary
• Reaching the oceanic basement at 5.5-6 km below
  the seafloor

52
Cold seeps on Nankai margin
Mud volcano
Splay fault
Kodaiba fault termination
Nautile (Ifremer) and Shinkai (Jamstec) manned
submersibles, ROVs and camera tracks (1985-2002).
53
Detecting anomalies in bottom water
54
Nankai cold seep dilemma
Shallow source hypothesis (footwall compaction)
Deep connection hypothesis (flow channelling
along thrusts)
55
Relationship between Marmara Sea
microseismicityand Ganos (1912) and Kocaeli
(1999) rupture zones
Kocaeli 1999
Ganos 1912
Maramara and Ganos fault systems (Le Pichon et
al., 2001 Tuyusuz et al. ,1998),
microearthquakes (Gurbuz et al., 2000), and GPS
vectors with respect to Eurasia (Meade et al.,
2002). Contours VIII and IX are seismic
intensity contours from Ganos 1912 earthquake
(Ambraseys and Finkel, 1987). According to these
authors, the IX contour limits the zone of slip.
56
Marmara Sea

•  Following the Kocaeli earthquake, many gas
  seepages have been observed on the sea bed. The
  most prominent of them was one about one km NW of
  Topçular where significant gas bubbles have been
  observed on the sea surface.  
•  Gas plumes are observed on the echo-sounder.
  They are sometimes very thick and diverted from
  vertical by currents. 
•  It is well-known that the subaquaeous
  sedimentary units in the Izmit Bay are charged
  with gas and the Holocene postransgressional
  marine deposits act as a cap layer. The
  gas-charged sediments are generally placed in the
  central parts of the basins where the gulf
  becomes wider. 
• From Bedri Alpar, Turkish Journal of Marine
  Sciences, 5 (3), 1999, "Underwater signatures of
  the Kocaeli earthquake,(August 17th 1999)",
  111-130. http//www.geocities.com/CapeCanaveral/St
  ation/8361/Quake99Turkey.html

57
Marmara seeps
Meteor cruise M44/1
bacterial mats along fault trace on western high
piston cores with shallow sulfate reduction zone
(5m)
methane in seawater
Marmascarps
widespread occurrence of black reduced patches
and bacterial mats
Inactive chimneys
Active chimneys
Active chimneys
58
Gas in sediment
T14
Jack the Smokerexpulsion of brackish fluid
Courtesy of R. Armijo and J. Malavieille,
Marmascarps co-chiefs
59
Marion Dufresne long cores (May 2004)

• Observations
• Gassy cores (3m geiser from holes in core liner)
• Shallow methane-sulfate reaction zone (4-8 m)
• Strong resistivity gradient indicative of
  brackish pore fluid
• Coarse sand layer gt12 m thick at 12.5 m below
  seafloor near chimney site
• Charging with gas and buoyancy drives fluid out
  of the sand layers through the seafloor fault
  rupture

cruise report http//cdf.u-3mrs.fr/henry/marmar
a
60
Cold seeps and active faults sum up

• Shallow process are dominant
• Methane bubble and water emissions may not occur
  at the same time nor at the same place
• Deeper sources may be present. However, fluid
  migration from the seismogenic zone has not been
  demonstrated for Nankai and Marmara Sea
• Even if source is shallow, relationship with
  earthquake cycle is likely from
• Coseismic fluidization
• Coseismic fracture opening
• Permeability change from continuous strain and
  stress change

61
EGS 2003

• Nankai cold seeps

62
Splay Faults and Cold Seeps on Nankai Margin

• P. Henry CNRS UMR 8538, ENS, Paris
• J. Ashi ORI, Univ. Tokyo
• U. Tsunogai, T. Toki Hokkaiko Univ., Sapporo
• S. Kuramoto, M. Kinoshita JAMSTEC, Yokosuka
• S.J. Lallemant Univ. Cergy-Pontoise
• C. Pierre LODYC, Paris 6
• Published work
• KAIKO-TOKAI Marine Geology special issue (2002)
• On going project
• SFJ (SEIZE Franco-Japonais) 3D seismics (Posters)
• Future project
• NanTroSEIZE (IODP) Deep (riser) drilling

63
Nankai Margin
To-Nankai 1944
Nankaido 1946 rupture
64
Cold seeps on Nankai margin
Mud volcano
Splay Fault
Kodaiba fault termination
Nautile (Ifremer) and Shinkai (Jamstec) manned
submersibles, ROVs and camera tracks (1985-2002).
65
Tokai Seep
Shallow source hypothesis footwall compaction
Nankai cold seep dilemma
Deep connection hypothesis flow channelling
along thrusts
66
Hydrate BSR Dome (High heat flow anomaly)on
Kodaiba Fault
67
DainiAtsumiKnollTermination of Kodaiba fault
Calyptogena sp. Carbonate crusts
Bathymodiolus sp.
Kodaiba fault
68
Kodaiba fault -gt Mud volcanoes
Kodaiba Fault
69
Mud volcanoes
70
Splay fault system
Line 4
71
Canyon system
72
Splay fault and accretionary wedgeLocation of
seeps on Line 4
73
Methane and Carbonates
Diagenetic cement formed at depths
Seawater C02 with Methane influence
Thermogenic methane (d13C -40 to -50)
Seawater C02
Biogenic methane (d13C lt -60) Seawater C02
KAIKO-NANKAI (Sakai et al., 1992)
74
Summary

• Major faults are active as conduits but
• Active seepage concentrates on short segments of
  fault outcrops (e.g. terminations)
• A large part of the fluid is derived from less
  than 1 km below sea floor (compaction biogenic
  methane)
• The most spectacular cold seeps result from
  thermogenic methane generation at depths,
  combined with active deformation

75
AGU Fall meeting 2003

• SFJ results on Tokai margin

76
Structure and evolution of a splay fault system
in the Tokai segment of Nankai Trough

• S.J. Lallemant Univ. Cergy-Pontoise
• P. Henry CNRS/Collège de France,
  Aix-en-Provence
• V. Martin ENS, Laboratoire de Géologie, Paris
• M. Noble Ecole des Mines de Paris, Fontainebleau
• J. X. Dessa CNRS,Géosciences Azur,
  Villefrance-sur-Mer
• S. Operto CNRS,Géosciences Azur,
  Villefrance-sur-Mer
• Present a summary of SFJ-Kaiko results
• Locate subducting basement topography
• Determine inner margin structure (backstop)
• Image splay faults - inactive and active

77
Map ofE Nankai

• Tokai segment
• Subducting basement ridge
• Oblique subduction
• Uplifted and deforming forearc basin
• Age constraints
• Submarine dive samples
• MITI drillholes

To-Nankai 1944

• SFJ survey
• MCS lines and 3D box
• shallow 3D prestack migration
• deep 2D prestack migration
• Dense OBS array
• Travel time tomography
• Waveform inversion

Nantroseize
78
Margin structure andsubduction megathrust
geometry 
Forearc basin
Active splays
Quaternary wedge
Pre-Miocene wedge
Paleo-Zenisu
top of crust
Paleo2-Zenisu ?
79
Subducting volcano movie
80
Blind splay faults within backstop
81
Active splay faults at backstop edge
Kodaiba
Tokai
82
Waveform inversion
Waveform inversion
Zenisu
paleo-zenisu
(paleo)2 -zenisu ?
83
Tokai thrust is a low velocity zone
84
Kodaiba fault and forearc basin history
Pliocene
-Gentle folding of the forearc basin occured in
two phases over the last 2 Ma -NW prograding
sequence associated with uplift above Kodaiba -On
this section Kodaiba is active as a propagating
blind fault
Early Miocene
Accretionary wedge (Eocene)
Ages from MITI drill hole
85
Kinematics

• Tokai segment has oblique subduction but no major
  shear partitionning fault on land.
• Splay faults (particularly Enshu F.) accomodate a
  significant right-lateral component.
• Subduction velocity and obliquity in Tokai region
  depends on the deformation of Central Japan and
  may vary in time

86
Conclusion

• Main Paleo-zenisu ridge is outside of the
  seismogenic zone in Tokai segment
• Backstop deformation occured during the Miocene
  and in two phases during the Quaternary
• Active splay faults have poor reflectivity but
  correspond to low velocity zones
• Splay fault activity depends on
• accretion of a frontal wedge during the
  quaternary
• subduction kinematics and shear partitionning
• subduction of basement topography