![Page 1: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/1.jpg)
W I T
Tomographic velocity model estimationwith data-derived first and second
spatial traveltime derivatives
Eric Duveneck, Tilman Klüver, Jürgen Mann∗
Geophysical InstituteUniversity of Karlsruhe
Germany
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 2: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/2.jpg)
W I TOverview
Introduction
Velocity determination with CRS attributes
A synthetic data example
A real data example
Extension to 3D
Advantages/Limitations
Conclusions
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 3: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/3.jpg)
W I TIntroduction
Problem: Determination of velocity modelfor depth imaging
Tomographic approach based on CRS stack results
Smooth model description
Advantages:picking in simulated ZO section of high S/N ratiopick locations independent of each other⇒ very few picks required
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 4: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/4.jpg)
W I TIntroduction
Problem: Determination of velocity modelfor depth imaging
Tomographic approach based on CRS stack results
Smooth model description
Advantages:picking in simulated ZO section of high S/N ratiopick locations independent of each other⇒ very few picks required
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 5: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/5.jpg)
W I TIntroduction
Problem: Determination of velocity modelfor depth imaging
Tomographic approach based on CRS stack results
Smooth model description
Advantages:picking in simulated ZO section of high S/N ratiopick locations independent of each other⇒ very few picks required
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 6: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/6.jpg)
W I TIntroduction
Problem: Determination of velocity modelfor depth imaging
Tomographic approach based on CRS stack results
Smooth model description
Advantages:picking in simulated ZO section of high S/N ratiopick locations independent of each other⇒ very few picks required
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 7: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/7.jpg)
W I TCRS stack – 3D data example
4000 −
2000 −
6000 −
8000 −
0 −
4000 −
2000 −
6000 −
8000 −
0 −
PreSDM PostSDM of CRS stack
Data courtesy of ENI E&P Division
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 8: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/8.jpg)
W I TCRS stack and attributes
t2 (ξm,h) =(
t0 + 2 sinαv0
(ξm−ξ ))2
+2 t0 cos2 α
v0
((ξm−ξ )2
RN+ h2
RNIP
)
ξ ξ
α α
NIP NIP
NIP NRR
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 9: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/9.jpg)
W I TCRS attributes and velocities
NIP
α
ξNIPR In the vicinity of a ZO ray:CRP-response can beapproximately described byt0, ξ , RNIP, α
Velocity model is consistentif RNIP = 0 at t = 0 for all con-sidered data points
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 10: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/10.jpg)
W I TCRS attributes and velocities
NIP
α
ξNIPR In the vicinity of a ZO ray:CRP-response can beapproximately described byt0, ξ , RNIP, α
Velocity model is consistentif RNIP = 0 at t = 0 for all con-sidered data points
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 11: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/11.jpg)
W I TTomography with CRS attributes
Data and model components
θ
ξ
α
MT
(x,z)
v(x,z)
Data: (T , M, α , ξ )i
Model: (x, z, θ )i, v jk
M = 1/v0RNIP
T = t0/2
v jk: B-spline coefficients
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 12: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/12.jpg)
W I TTomography with CRS attributes
Data and model components
θ
ξ
α
MT
(x,z)
v(x,z)
Data: (T , M, α , ξ )i
Model: (x, z, θ )i, v jk
M = 1/v0RNIP
T = t0/2
v jk: B-spline coefficients
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 13: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/13.jpg)
W I TForward modeling
Kinematic ray-tracing
⇒ T , α , ξ
Dynamic ray-tracing
⇒ Ray propagator matrix ΠΠΠ =
(Q1 Q2P1 P2
)
⇒ M = P2/Q2
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 14: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/14.jpg)
W I TForward modeling
Kinematic ray-tracing
⇒ T , α , ξ
Dynamic ray-tracing
⇒ Ray propagator matrix ΠΠΠ =
(Q1 Q2P1 P2
)
⇒ M = P2/Q2
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 15: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/15.jpg)
W I TInversion procedure
nonlinear least-squares problem
⇒ iterative solution, linearize locally
model update ∆m: least-squares solution of
F∆m = ∆d
with ∆d : data misfitF : Fréchet derivatives
calculation of Fréchet derivatives:ray perturbation theory
regularization ⇒ F̂∆m = ∆d̂(minimization of second derivatives of velocity)
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 16: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/16.jpg)
W I TInversion procedure
nonlinear least-squares problem
⇒ iterative solution, linearize locally
model update ∆m: least-squares solution of
F∆m = ∆d
with ∆d : data misfitF : Fréchet derivatives
calculation of Fréchet derivatives:ray perturbation theory
regularization ⇒ F̂∆m = ∆d̂(minimization of second derivatives of velocity)
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 17: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/17.jpg)
W I TInversion procedure
nonlinear least-squares problem
⇒ iterative solution, linearize locally
model update ∆m: least-squares solution of
F∆m = ∆d
with ∆d : data misfitF : Fréchet derivatives
calculation of Fréchet derivatives:ray perturbation theory
regularization ⇒ F̂∆m = ∆d̂(minimization of second derivatives of velocity)
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 18: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/18.jpg)
W I TInversion procedure
nonlinear least-squares problem
⇒ iterative solution, linearize locally
model update ∆m: least-squares solution of
F∆m = ∆d
with ∆d : data misfitF : Fréchet derivatives
calculation of Fréchet derivatives:ray perturbation theory
regularization ⇒ F̂∆m = ∆d̂(minimization of second derivatives of velocity)
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 19: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/19.jpg)
W I TA synthetic data example
Original velocity model
-1500 500 2500 4500 6500x [m]
-3000
-2000
-1000
0
z [m
]
2000
3000
4000
5000
m/s
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 20: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/20.jpg)
W I TSynthetic data example
0
1
2
t [s]
0 2 4 6x [km]
0
1
2
t [s]
0 2 4 6x [km]
0
5
10
Rni
p [k
m]
0
1
2
t [s]
0 2 4 6x [km]
-20
0
20
angl
e [°
]
CRS stack RNIP section α section
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 21: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/21.jpg)
W I TSynthetic data example
Picked input data for the inversion
−2000 0 2000 4000 6000 8000
0
200
400
600
800
1000
1200
1400
Xo [m]
T [1
0−3
s]
−2000 0 2000 4000 6000 80000
200
400
600
800
1000
1200
1400
1600
1800
Xo [m]
M [1
0−9
s/m
2 ]
−2000 0 2000 4000 6000 8000−30
−20
−10
0
10
20
30
Xo [m]
alph
a [
o ]
T M α
Model parametrization:B-spline knot spacing ∆x = 500 m, ∆z = 300 m
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 22: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/22.jpg)
W I TSynthetic data example
Residual data error after 12 iterations
−2000 0 2000 4000 6000 8000−6
−4
−2
0
2
4
6
Xo [m]
Del
ta T
[10
−3 s
]
6
4
2
0
−2
−4−6
−2000 0 2000 4000 6000 8000−8
−6
−4
−2
0
2
4
6
8
10
Xo [m]
Del
ta M
[10
−9 s
/m2 ]
8
4
0
−4
−8−2000 0 2000 4000 6000 8000
−0.02
−0.015
−0.01
−0.005
0
0.005
0.01
0.015
0.02
0.025
Xo [m]
Del
ta a
lpha
[ o ]
0.02
0.01
0
−0.01
−0.02
∆T [10−3s] ∆M [10−9 s/m2] ∆α [◦]
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 23: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/23.jpg)
W I TSynthetic data example
Inversion result
-1500 500 2500 4500 6500x [m]
-3000
-2000
-1000
0
z [m
]
2000
3000
4000
5000
m/s
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 24: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/24.jpg)
W I TSynthetic data example
Inversion result
-1500 500 2500 4500 6500x [m]
-3000
-2000
-1000
0
z [m
]
2000
3000
4000
5000
m/s
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 25: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/25.jpg)
W I TReconstructed vs. original model
Reconstructed velocity and reflector elements
-1500 500 2500 4500 6500x [m]
-3000
-2000
-1000
0
z [m
]
2000
3000
4000
5000
m/s
Original velocity and reconstructed reflector elements
-1500 500 2500 4500 6500x [m]
-3000
-2000
-1000
0
z [m
]
2000
3000
4000
5000
m/s
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 26: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/26.jpg)
W I TPrestack migration results
1
2
3
z [k
m]
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0CIG location [km]
common-image gatherscommon-image gathers (maximum offset=2000m)
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 27: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/27.jpg)
W I TIncluding additional constraints
. . . important in the case of data gaps!
v(x,z) values at arbitrary locations (x,z)
spatially dependent regularization(smoothness of velocity model)
force velocity structure to follow local reflectorstructure
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 28: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/28.jpg)
W I TIncluding additional constraints
. . . important in the case of data gaps!
v(x,z) values at arbitrary locations (x,z)
spatially dependent regularization(smoothness of velocity model)
force velocity structure to follow local reflectorstructure
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 29: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/29.jpg)
W I TIncluding additional constraints
. . . important in the case of data gaps!
v(x,z) values at arbitrary locations (x,z)
spatially dependent regularization(smoothness of velocity model)
force velocity structure to follow local reflectorstructure
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 30: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/30.jpg)
W I TReal data example
0
1
2
3
t [s]
0 1 2 3 4 5 6x [km]
CRS stack
0
1
2
3t [
s]
0 1 2 3 4 5 6x [km]
coherence
0
0.1
0.2
0.3
0.4
0.5
CRS stack section coherence section(semblance)
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 31: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/31.jpg)
W I TReal data example
0
1
2
3
t [s]
0 1 2 3 4 5 6x [km]
Rnip
0
2
4
6
8
10
0
1
2
3t [
s]
0 1 2 3 4 5 6x [km]
angle
-40
-20
0
20
40
RNIP section [km] angle section [◦]
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 32: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/32.jpg)
W I TReal data example
Picked input data for the inversion
0 1000 2000 3000 4000 5000 6000
0
200
400
600
800
1000
1200
1400
x [m]
T [
10−3
s]
0 1000 2000 3000 4000 5000 60000
200
400
600
800
1000
1200
1400
1600
1800
x [m]
M [
10−9
s/m
2 ]
0 1000 2000 3000 4000 5000 6000−20
−15
−10
−5
0
5
x [m]
angl
e [
o ]
T M α
Model parametrization:B-spline knot spacing ∆x = 500 m, ∆z = 300 m
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 33: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/33.jpg)
W I TReal data example
Residual data error after 12 iterations
0 1000 2000 3000 4000 5000 6000−20
−15
−10
−5
0
5
10
15
x [m]
Del
ta T
[10
−3 s
]
0
10
−10
0 1000 2000 3000 4000 5000 6000−80
−60
−40
−20
0
20
40
60
x [m]
Del
ta M
[10
−9 s
/m2 ]
0
−40
40
0 1000 2000 3000 4000 5000 6000−0.2
−0.15
−0.1
−0.05
0
0.05
0.1
0.15
x [m]
Del
ta a
ngle
[ o ]
0
−0.1
0.1
∆T [10−3s] ∆M [10−9 s/m2] ∆α [◦]
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 34: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/34.jpg)
W I TReal data example
Inversion result0
1000
2000
3000
4000
5000
z [m
]0 1000 2000 3000 4000 5000 6000
x [m]
inversion result
2000
3000
4000
5000
velo
city
[m/s
]
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 35: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/35.jpg)
W I TReal data example
Inversion result0
1000
2000
3000
4000
5000
z [m
]0 1000 2000 3000 4000 5000 6000
x [m]
inversion result
2000
3000
4000
5000
velo
city
[m/s
]
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 36: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/36.jpg)
W I TPoststack migration
0
1000
2000
3000
4000
5000
z [m
]
0 1000 2000 3000 4000 5000 6000x [m]
inversion result
2000
3000
4000
5000
velo
city
[m/s
]
1
2
3
4
5
z [k
m]
0 1 2 3 4 5 6x [km]
CRS poststack migration
Reconstructed velocity Poststack migrationmodel and dip bars of CRS stack result
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 37: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/37.jpg)
W I TPrestack migration results
1
2
3
4
5
z [k
m]
1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0CIG location [km]
common-image gatherscommon-image gathers (maximum offset=2000m)
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 38: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/38.jpg)
W I T3D CRS attributes
t2 (∆ξξξ ,h) =(
t0 + 2pξ ·∆ξξξ)2
+2t0
(∆ξξξ T Mξ ∆ξξξ + hT Mh h
)
pξ = 12∂ t/∂ξξξ = 1
v0(sinα cosψ ,sinα sinψ)T
Mh = 12∂ 2t/∂h2 = 1
v0DKNIPDT
Mξ = 12∂ 2t/∂ξξξ 2 = 1
v0DKNDT
Independent of near-surface velocity v0 !8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 39: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/39.jpg)
W I T3D CRS attributes
t2 (∆ξξξ ,h) =(
t0 + 2pξ ·∆ξξξ)2
+2t0
(∆ξξξ T Mξ ∆ξξξ + hT Mh h
)
For a tomographic inversion, only
one azimuth φ of Mh is required: Mφ !
⇒ Data: (T ,Mφ , pξx, pξy
,ξx,ξy)
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 40: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/40.jpg)
W I T3D inversion with CRS attributes
Data and model components
(ξ ,ξ )
x
x y
y
yx φ
(x,y,z)
(e ,e )
M(p ,p )T
v(x,y,z)
Data:(T , Mφ , pξx
, pξy, ξx, ξy)i
Model:(x, y, z, ex, ey)i, v jkl
T = t0/2
v jkl: B-spline coefficients
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 41: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/41.jpg)
W I T3D inversion with CRS attributes
Data and model components
(ξ ,ξ )
x
x y
y
yx φ
(x,y,z)
(e ,e )
M(p ,p )T
v(x,y,z)
Data:(T , Mφ , pξx
, pξy, ξx, ξy)i
Model:(x, y, z, ex, ey)i, v jkl
T = t0/2
v jkl: B-spline coefficients
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 42: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/42.jpg)
W I T3D synthetic example
Cut through original and reconstructed 3D models
Model
v [k
m/s
]
5
43
2
34
5
x [km]
01
y [km]2
16
1.5
4.3
original model
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 43: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/43.jpg)
W I T3D synthetic example
Cut through original and reconstructed 3D models
Inversion result
v [k
m/s
]
5
43
2
34
5
x [km]
01
y [km]2
16
1.5
4.3
inversion result
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 44: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/44.jpg)
W I T3D synthetic example
Depth slice at z=1500 m
Model
0
1
2
3
4
5
1 2 3 4 5 6y [km]
4.3
z = 1500 m
x [km]
v [k
m/s
]
1.5
original model8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 45: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/45.jpg)
W I T3D synthetic example
Depth slice at z=1500 m
Inversion result
0
1
2
3
4
5
1 2 3 4 5 6y [km]
4.3
x [km]
v [k
m/s
]
z = 1500 m
1.5
inversion result8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 46: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/46.jpg)
W I TAdvantages/Limitations
Input is a by-product of CRS stack
Very few picks are required
Picking in ZO section of increased S/N ratio
No assumptions about reflector continuity
Smooth model (ideal for ray-tracing)
Smooth velocity description must be valid
Limited lateral variation within CRS aperture(approximately hyperbolic traveltimes)
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 47: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/47.jpg)
W I TAdvantages/Limitations
Input is a by-product of CRS stack
Very few picks are required
Picking in ZO section of increased S/N ratio
No assumptions about reflector continuity
Smooth model (ideal for ray-tracing)
Smooth velocity description must be valid
Limited lateral variation within CRS aperture(approximately hyperbolic traveltimes)
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 48: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/48.jpg)
W I TAdvantages/Limitations
Input is a by-product of CRS stack
Very few picks are required
Picking in ZO section of increased S/N ratio
No assumptions about reflector continuity
Smooth model (ideal for ray-tracing)
Smooth velocity description must be valid
Limited lateral variation within CRS aperture(approximately hyperbolic traveltimes)
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 49: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/49.jpg)
W I TAdvantages/Limitations
Input is a by-product of CRS stack
Very few picks are required
Picking in ZO section of increased S/N ratio
No assumptions about reflector continuity
Smooth model (ideal for ray-tracing)
Smooth velocity description must be valid
Limited lateral variation within CRS aperture(approximately hyperbolic traveltimes)
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 50: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/50.jpg)
W I TAdvantages/Limitations
Input is a by-product of CRS stack
Very few picks are required
Picking in ZO section of increased S/N ratio
No assumptions about reflector continuity
Smooth model (ideal for ray-tracing)
Smooth velocity description must be valid
Limited lateral variation within CRS aperture(approximately hyperbolic traveltimes)
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 51: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/51.jpg)
W I TAdvantages/Limitations
Input is a by-product of CRS stack
Very few picks are required
Picking in ZO section of increased S/N ratio
No assumptions about reflector continuity
Smooth model (ideal for ray-tracing)
Smooth velocity description must be valid
Limited lateral variation within CRS aperture(approximately hyperbolic traveltimes)
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 52: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/52.jpg)
W I TAdvantages/Limitations
Input is a by-product of CRS stack
Very few picks are required
Picking in ZO section of increased S/N ratio
No assumptions about reflector continuity
Smooth model (ideal for ray-tracing)
Smooth velocity description must be valid
Limited lateral variation within CRS aperture(approximately hyperbolic traveltimes)
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 53: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/53.jpg)
W I TAdvantages/Limitations
Input is a by-product of CRS stack
Very few picks are required
Picking in ZO section of increased S/N ratio
No assumptions about reflector continuity
Smooth model (ideal for ray-tracing)
Smooth velocity description must be valid
Limited lateral variation within CRS aperture(approximately hyperbolic traveltimes)
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 54: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/54.jpg)
W I TConclusions
CRS stack yields information useful fordetermination of smooth velocity modelsfor depth imaging
Tomographic inversion method based onCRS attributes
Implementation in 2D and 3D
Applied to 2D real data
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 55: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/55.jpg)
W I TConclusions
CRS stack yields information useful fordetermination of smooth velocity modelsfor depth imaging
Tomographic inversion method based onCRS attributes
Implementation in 2D and 3D
Applied to 2D real data
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 56: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/56.jpg)
W I TConclusions
CRS stack yields information useful fordetermination of smooth velocity modelsfor depth imaging
Tomographic inversion method based onCRS attributes
Implementation in 2D and 3D
Applied to 2D real data
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 57: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/57.jpg)
W I TConclusions
CRS stack yields information useful fordetermination of smooth velocity modelsfor depth imaging
Tomographic inversion method based onCRS attributes
Implementation in 2D and 3D
Applied to 2D real data
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 58: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/58.jpg)
W I TAcknowledgment
This work has been supported by the sponsors of theWave Inversion Technology (WIT) consortium.
W I T
8th Int. Congress, Brasilian Geophysical Society, Rio 2003
![Page 59: Tomographic velocity model estimation with data-derived ... · Tomographic velocity model estimation with data-derived first and second spatial traveltime derivatives Eric Duveneck,](https://reader033.vdocument.in/reader033/viewer/2022041418/5e1d23e7c3c8f529c77dfebd/html5/thumbnails/59.jpg)
W I T
8th Int. Congress, Brasilian Geophysical Society, Rio 2003