interseismic deformation with aseismic stress-dependent fault slip

1
interseismic deformation with aseismic
stress-dependent fault slip
a very informal, and preliminary talk about how
we are thinking about

• Eric A Hetland, Mark Simons, Ravi Kanda, Sue Owen

TO brown-bag 03 April 2007
2
post-seismic slip following subduction ruptures
2005 Nias-Simeulue eq. (M8.7)
fault rheology is not (explicitly) included in
after-slip model
3
post-seismic slip following subduction ruptures
1995 Antofagasta eq. (M8.1)
fault rheology is not (explicitly) included in
after-slip model
4
post-seismic slip following subduction ruptures
2003 Tokachi-oki eq. (M8)
fault rheology is not included in after-slip model
5
inter-seismic slip near regions of past
subduction ruptures
Japan/southern Kurile trenches
Suwa et al., 2006
model assumes fault slip during inter-seismic
period is constant
6
we want an internally consistent model that can
describe observations of both inter-seismic and
post-seismic deformation
for now we are building subduction zone models
that include repeated ruptures, on assumed
asperities, with stress-dependent aseismic slip
on the non-asperity portions of the subduction
interface during the interseismic period
7
(No Transcript)
8
(No Transcript)
9
introduced by J. Savage (Savage and Burford,
1973 Savage and Prescott, 1978 Savage, 1983) as
a mathematically convenient fault loading
mechanism in kinematic quasi-kinematic models
approximation only good for spun-up systems rate
of interseismic relaxation rate of reloading
10
we impose ruptures - we do not solve for them
11
we impose ruptures - we do not solve for them
imposed ruptures at times Tp
long-term fault-slip
interseismic slip on fault
traction on fault
part of fault with coseismic slip
part of fault that slips steadily
part of fault that is allowed to slip
interseismically
need a fault rheology
12
rate- and state-friction
(a-b)lt0 ? ruptures, (a-b)gt0 ? aseismic slip ?
is a state variable, assume it is constant ? ?
L/v ? ? ?N
Dieterich, 1979 Ruina 1983 Rice and Gu, 1983
(figure from Ben-Zion, 2003)
13
(No Transcript)
14
(No Transcript)
15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
we impose ruptures - we only solve for aseismic
slip
fault rheology
bulk rheology
for now, assume elastic half-space and use Okada,
1992
use boundary elements
model works for 3D, non-planar faults, with
multiple asperities, arbitrary rheologic
parameters, we allow both dip- and strike-slip
co- and inter-seismic slip, and irregular
(imposed) rupture sequences
currently, we can impose coseismic slip in
non-locked regions of the fault, but we do not
allow interseismic slip in the locked regions
19
? 30 GPa ?N 300 MPa D 104 m bo 10 m ?
(a-b) ?-1/10 ?-1 0.5 ? (a-b) 0.05 ?-1 1.0
? (a-b) 0.10
20
thrust fault in an elastic half-space, dipping
45 degrees
21
thrust fault in an elastic half-space, dipping
45 degrees
a more realistic geometry
back-slip model
interseismic surface deformation is given by the
locked portions of the mega-thrust sliding as a
normal fault at the plate rate (Savage, 1983)
vertical horizontal
22
thrust fault in an elastic half-space, dipping
45 degrees
a more realistic geometry
elastic slab model
does not include strains due to plate bending, if
incorporated, discrepancy removed, total
interseismic coseismic subduction block
motion
vertical horizontal
Ravi Kanda
23
thrust fault in an elastic half-space, dipping
45 degrees
periodically impose this co-seismic slip
24
slip on the fault
below the locked region
bgt0 ? thrust slip
25
surface interseismic displacements
26
surface interseismic displacements
27
surface interseismic displacements motivation
28
surface interseismic displacements motivation
x
29
determination of plate coupling
shown is back-slip rate vbs
this assumes that the interseismic deformation is
constant throughout the interseismic period

• invert GPS velocities for distributions of normal
  slip (vbs) on the mega-thrust
• use back-slip model (Savage, 1983) to determine
  the coupling coefficient
• vbs vT ? coupled (C1)
• vbs 0 ? uncoupled (C0)

30
determination of plate coupling
this assumes that the interseismic deformation is
constant throughout the interseismic period

• invert GPS velocities for distributions of normal
  slip (vbs) on the mega-thrust
• use back-slip model (Savage, 1983) to determine
  the coupling coefficient
• vbs vT ? coupled (C1)
• vbs 0 ? uncoupled (C0)

slip is not constant through the cycle
31
variation of coupling through an interseismic
period
xxxxxx
32
variation of coupling through an interseismic
period
33
variation of coupling through an interseismic
period
34
this model only contains co-seismic slip in the
locked regions, no interseismic slip-allowed in
the locked regions contrary to dynamic
calculations
35
two (of the many) remaining issues still
learning to drive lockedness we assume full
slip in locked patches (asperities) some
directions currently aiming for include
heterogeneous elastic structure by computing
K(z?) from FE models include other bulk
rheologies K(z?) simple semi-analytic
models quite complicated FE models model the
GPS data of inter- post-seismic observations in
Hokkaido (2D, 3D planar, respecting slab
geometry, )
36
gOcad
1973
2003
1968
slip models from Yamanaka and Kikuchi (2002)
vertically exaggerated