Geodesy and Geodynamics
By
Christophe Vigny
National Center for scientific Research (CNRS)
&
Ecole Normale Supérieure (ENS) Paris, France
http://www.geologie.ens.fr/~vigny
DEFORMATION PATTERN IN ELASTIC CRUST
•
Stress and force in 2D
•
Strain : normal and shear
•
Elastic medium equations
•
Vertical fault in elastic medium => arctangent
•
General elastic dislocation (Okada’s formulas)
•
Fault examples
Stress ( ) in 2D
- Force =
xsurface
- no rotation =>
xy=
- only 3 independent
yx….components :
…..
xx,
yy,
xy- Normal stress = ii
- Shear stress = ij
Applied forces
Normal forces on x axis
xx(x).
y
xx(x+
x).
y
y
xx(x).
xx(x+
x)
y d
xx/d
x .
x (1) Shear forces on x axis
yx(y).
x
yx(y+
y).
x
x d
yx/d
y .
y (2) Total on x axis
d
xx/d
x + d
yx/d
y
x
yForces Equilibrium
Total on x axis =
d
xx/d
x + d
yx/d
y
x
yTotal on y axis =
d
yy/d
y + d
yx/d
x
y
x d
yy/d
y + d
yx/d
x
d
xx/d
x + d
yx/d
y
Equilibrium =>
Normal strain : change length (not angles)
- Change of length proportional to length
-
xx,
yy,
zz are normal component of strain
+
+
Shear strain : change angles
xy = -1/2(
1 +
2)
= 1/2(
dy
dx + dx
dy)
xy =
yx (obvious)Solid elastic deformation (1)
• Stresses are proportional to strains
• No preferred orientations
xx = ( G)
xx +
yy +
zz
yy =
xx + ( G)
yy +
zz
zz =
xx +
yy + ( G)
zz• andG are Lamé parameters
The material properties are such that a principal strain component
produces a stress ( G ) in the same direction
and stresses in mutually perpendicular directions
Solid elastic deformation (2)
Inversing stresses and strains give :
xx = 1/
E
xx - /
E
yy - /
E
zz
yy = - /
E
xx + 1/
E
yy - /
E
zz
zz = - /
E
xx - /
E
yy + 1/
E
zz• E and are Young’s modulus and Poisson ’s ratio
a principal stress component produces
a strain
1/
E in the same direction and
strains / in mutually perpendicular directions
III/ Elastic deformation at plate boundaries
What is the shape of the accumulated deformation ?
Two plates (red and green) are separated by a vertical strike-slip fault.
If the fault was slick then the two plates would slide freely along each other with no deformation. But because its surface is rough and there is some friction, the fault is locked. So because the plates keep moving far away from the fault, and don’t move on the fault, they have to deform.
Mathematical formulation
The plates are made of elastic crust above a viscous mantle.
The viscous flow localizes the deformation in a narrow band just beneath the elastic layer
We can compute the deformation of the elastic layer using the elastic equations detailed before
Mathematical formulation
•Symetry => all derivative with y = 0
yy= 0
•No gravity =>
zz= 0
•What is the displacement field U in the elastic layer ?
Mathematical formulation
•Elastic equations :
(3)
+
zz = 0 =>
xx +
zz = -2G
zzand
(2) =>
yy =
xx +
zz = -2G
zz
xx = ( G)
xx +
zz(2)
yy =
xx +
zz(3)
zz =
xx +(
G)
zz
xy =G
xy
xz =G
xz
yz =G
yz=>
xx = - (2G
+
zzand
(1) =>
xx =[
- (2G)2/ + ]
zz_________________________
Mathematical formulation
• Force equilibrium along the 3 axis
(x)
d
xx/ d x
+d
yx/ d
y +d
xz/ d z
= 0_________________________
(y)
d
xy/ d x
+d
yy/ d
y +d
yz/ d z
= 0(z)
d
xz/ d x
+d
yz/ d
y +d
zz/ d z
= 0x x
x x
• Derivation of eq. 1 with x and eq. 3 give :
d 2
xx/ d x 2
= 0• equation 2 becomes :
d
xy/ d x
+d
yz/ d z
= 0Mathematical formulation
relations between stress (
) and displacement vector (U)
xy = 2G
xy = 2G [ d U
x/ d
y +d U
y/ d x ]
.1/
2_________________________
yz = 2G
yz = 2G [ d U
z/ d
y +d U
y/ d z ]
.1/
2x
d / d x [ d U
x/ d
y +d U
y/ d x ]
+d / d z [ d U
z/ d
y +d U
y/ d z ]
= 0d 2 U
y/ d x 2
+d 2 U
y/ d z 2
= 0Using
d
xy/ d x
+d
yz/ d z
= 0 we obtain :x
Mathematical formulation
What is
U
y, function of x and z, solution of this equation ?d 2 U
y/ d x 2
+d 2 U
y/ d z 2
= 0Guess :
U
y = K arctang (x/
z) works fine !Nb.
d
atan(
)/ d
1/
(1+2 )d U
y/ d
x=
K/
z(1+x2/z2)=> d
2U
y/ d
x2= -2Kxz/
(z2+x2)d U
y/ d
z=
-Kx/
z2 (1+x2/z2)=> d
2U
y/ d
z2= 2Kxz/
(x2+z2)Mathematical formulation
Boundary condition at the base of the crust (z=0)
U
y = K arctang (x/
z)U
y = K . /2 if x > 0 = K . – /2 if x < 0=> K = 2.
V0/
And also :
U
y = +V0 if x > 0 = –V0 if x < 0Mathematical formulation
at the surface (z=h)
U
y = K arctang (x/
z)U
y =2.
V0/
arctang (x/
h)The expected profile of deformation across a strike slip fault we should see at the surface of the earth (if the crust is elastic) is shape like an arctangant function. The exact shape depends on the thickness of the elastic crust, also
Arctang profiles U
y =2.
V0/
arctang (x/
h)Sagaing Fault, Myanmar
GPS measurement on the Sagaing fault fit well the arctang profile
but with an offset of 10- 15 km
Palu Fault, Sulawesi
Part of the GPS data on Palu fault fits well an arctang profile. But wee need a second fault to explain all the
Altyn Tagh Fault, China
Altyn Tagh Fault, China (INSAR)
Interferogram Nov. 1995/ Nov. 1999
Fault-parallel velocity :
v = v0/ atan(x/D)
Slip rate V0 = 1.4 cm/yr Locking depth D = 15 km
San Andreas Fault, USA (INSAR)
Elastic dislocation (Okada, 1985)
The displacement field ui(x1,x2,x3) due to a dislocation uj (1,2,3) across a surface in an isotropic medium is given by :
Where jk is the Kronecker delta, and are Lamé’s parameters, k is the direction cosine of the normal to the surface element d.
uij is the ith component of the
displacement at (x1,x2,x3) due to the jth direction point force of magnitude
Surface deformation due to shear and tensile faults in a half space, BSSA vol75, n°4, 1135-1154, 1985.
Elastic dislocation (Okada, 1985)
(1) displacements
For strike-slip For dip-slip For tensile fault
Elastic dislocation (Okada, 1985)
Where :
Subduction modeling
Velocity component // to convergence direction In the case of a subduction (dippping fault with downward slip) we use Okada’s formulas.
We find a very large deformation area (> 500 km) because the dipping angle is only 22°
With oblique slip we predict the surface vector will start to rotate above the end-tip of the subduction plane The profile of the velocity component // to the convergence shows a
“plateau” at this location
Subduction elastic curves
Normalized horizontal and vertical curves : . The applied dislocation on the subduction plane is equivalent to 1 cm/yr accumulation
GPS measurements in Chile and S-America
“fiducial” stations on the stable South American craton (Brazil, Guyana, Argentina, Urugay,) allow to define the:
South-American reference frame.
GPS measurements in Chile
Deformation (elastic def. induced by coupling on the subduction) is visible in Chile (Constitucion)
And
reaches far inland:
TUCU (Tucuman) and CFAG (Coronel
Fontana) in
Argentina show deformation more than 400 km away from the trench
Velocity profiles
As expected from elastic coupling, velocities decrease Eastward (from 35- 45 mm/yr along the coast to 10-15 mm/yr at the cordillera) and vector directions rotate from a direction // to plate convergence to East-West trending.
Subduction parameter adjustments
Elastic curves fit
Along strike comparison
Conception 37°S Coquimbo 30°S
Along strike comparison
Patial coupling model
potential for an Eq potential for an Eq . . Mw ~8.5 every Mw ~8.5 every plane: dip angle 15°
plane: dip angle 15°
. . locked at locked at 50km50km
slip: 2 – 2.5 cm/an slip: 2 – 2.5 cm/an
.. N 30°-35° N 30°-35°