3-D PRESTACK WAVEPATH MIGRATION H. Sun Geology and Geophysics Department University of Utah.

3-D PRESTACK WAVEPATH 3-D PRESTACK WAVEPATH MIGRATIONMIGRATION

H. SunH. Sun

Geology and Geophysics Department Geology and Geophysics Department University of UtahUniversity of Utah

OutlineOutline

Problems in 2-D WMProblems in 2-D WM• Objectives of 3-D WM Objectives of 3-D WM • Numerical ResultsNumerical Results• ConclusionsConclusions

• Lower CPU EfficiencyLower CPU Efficiency only 1/3 faster than KMonly 1/3 faster than KM or even 2 times slower than KMor even 2 times slower than KM• Moderate CPU EfficiencyModerate CPU Efficiency 4~11 times faster than KM4~11 times faster than KM By Slant StackingBy Slant Stacking

Problems in 2-D WMProblems in 2-D WM

2-D KM of a Single Trace2-D KM of a Single Trace

RR SS

NNNN

11

11CPU Count = NCPU Count = N 22

AA

AA

BB

BB

CC

CC

2-D WM of a Single Trace2-D WM of a Single Trace

RR SSAA

BBCC

CPU Count = 3 * ( Tracing + Searching + Migrating )CPU Count = 3 * ( Tracing + Searching + Migrating ) = 3 * ( N + 2N + 7N) = 30N = 3 * ( N + 2N + 7N) = 30N

NNNN

11

11

AABB

CC

OutlineOutline

• Problems in 2-D WMProblems in 2-D WM Objectives of 3-D WM Objectives of 3-D WM • Numerical ResultsNumerical Results• ConclusionsConclusions

• To Achieve Higher CPU EfficiencyTo Achieve Higher CPU Efficiency Compared to 3-D KMCompared to 3-D KM • To Generate Comparable or BetterTo Generate Comparable or Better Image Quality than 3-D KMImage Quality than 3-D KM

Key Goals of 3-D WMKey Goals of 3-D WM

OutlineOutline• Problems in 2-D WMProblems in 2-D WM• Objectives of 3-D WM Objectives of 3-D WM Numerical ResultsNumerical Results 3-D Point Scatterer Data3-D Point Scatterer Data 3-D SEG/EAGE Salt Data3-D SEG/EAGE Salt Data 3-D West Texas Field Data3-D West Texas Field Data• ConclusionsConclusions

3-D Prestack KM Point Scatterer Response3-D Prestack KM Point Scatterer Response R

efle

ctiv

ity

Ref

lect

ivit

y

Y Offset (km)Y Offset (km) X Offset (km)X Offset (km)

11

-0.5-0.5

00

11

Ref

lect

ivit

yR

efle

ctiv

ity

Y Offset (km)Y Offset (km) X Offset (km)X Offset (km)

11

-0.01-0.01

00

0.020.02

Ref

lect

ivit

yR

efle

ctiv

ity

Y Offset (km)Y Offset (km) X Offset (km)X Offset (km)

11

-0.05-0.05

00

0.10.1

Ref

lect

ivit

yR

efle

ctiv

ity

Y Offset (km)Y Offset (km) X Offset (km)X Offset (km)

11

-0.2-0.2

00

0.40.4

11

1111

11

Z0Z0

Z0-1Z0-1Z0-9Z0-9

Z0+8Z0+8

Ref

lect

ivit

yR

efle

ctiv

ity

Y Offset (km)Y Offset (km) X Offset (km)X Offset (km)

11

-0.5-0.5

00

11

Ref

lect

ivit

yR

efle

ctiv

ity

Y Offset (km)Y Offset (km) X Offset (km)X Offset (km)

11

-0.01-0.01

00

0.020.02

Ref

lect

ivit

yR

efle

ctiv

ity

Y Offset (km)Y Offset (km) X Offset (km)X Offset (km)

11

-0.05-0.05

00

0.10.1

Ref

lect

ivit

yR

efle

ctiv

ity

Y Offset (km)Y Offset (km) X Offset (km)X Offset (km)

11

-0.2-0.2

00

0.40.4

11

1111

11

3-D Prestack WM Point Scatterer Response3-D Prestack WM Point Scatterer Response

Z0Z0

Z0-1Z0-1Z0-9Z0-9

Z0+8Z0+8

OutlineOutline• Problems in 2-D WMProblems in 2-D WM• Objectives of 3-D WM Objectives of 3-D WM Numerical ResultsNumerical Results 3-D Point Scatterer Data3-D Point Scatterer Data 3-D SEG/EAGE Salt Data3-D SEG/EAGE Salt Data 3-D West Texas Field Data3-D West Texas Field Data• ConclusionsConclusions

A Common Shot GatherA Common Shot GatherTrace NumberTrace Number11 390390

Tim

e (s

ec)

Tim

e (s

ec)

00

5.05.0

Receiver DistributionReceiver DistributionC

ross

line

(m)

Cro

sslin

e (m

)

44804480

23202320

19201920

19201920 Inline (m)Inline (m)

Inline Velocity ModelInline Velocity Model

Offset (km)Offset (km)00 9.29.2

Dep

th (

km)

Dep

th (

km)

00

3.83.8

SALTSALT

Inline KMInline KM ((CPU=1CPU=1)) Inline WMInline WM ((CPU=1/33CPU=1/33))

Offset (km)Offset (km)00 9.29.2

00

3.83.8

De

pth

(k

m)

De

pth

(k

m)

Offset (km)Offset (km)00 9.29.2

Receiver DistributionReceiver DistributionC

ross

line

(m)

Cro

sslin

e (m

)

44804480

23202320

19201920

19201920 Inline (m)Inline (m)

Inline KMInline KM ((CPU=1CPU=1)) Inline WMInline WM ((CPU=1/170CPU=1/170))

Offset (km)Offset (km)00 9.29.2

00

3.83.8

De

pth

(k

m)

De

pth

(k

m)

Offset (km)Offset (km)00 9.29.2

(subsample)(subsample)

Zoom Views of Inline Sections Zoom Views of Inline Sections

Offset: 3~6.5 km, Depth: 0.3~1.8 kmOffset: 3~6.5 km, Depth: 0.3~1.8 km

WMWM

ModelModel

KM KM

SubSubWMWM

Offset: 1.8~4 km, Depth: 0.6~2.1 kmOffset: 1.8~4 km, Depth: 0.6~2.1 km

WMWM

ModelModel

KM KM

SubSubWMWM

Zoom Views of Crossline Sections Zoom Views of Crossline Sections

Inline: 1.8~7.2 km, Crossline: 0~4 kmInline: 1.8~7.2 km, Crossline: 0~4 km

WMWM

ModelModel

KM KM

SubSubWMWM

Horizontal Slices (Depth=1.4 km) Horizontal Slices (Depth=1.4 km)

OutlineOutline• Problems in 2-D WMProblems in 2-D WM• Objectives of 3-D WM Objectives of 3-D WM Numerical ResultsNumerical Results 3-D Point Scatterer Data3-D Point Scatterer Data 3-D SEG/EAGE Salt Data3-D SEG/EAGE Salt Data 3-D West Texas Field Data3-D West Texas Field Data• ConclusionsConclusions

A Common Shot GatherA Common Shot GatherTrace NumberTrace Number5454 193193

Tim

e (s

ec)

Tim

e (s

ec)

00

3.43.4

Receiver DistributionReceiver DistributionC

ross

line

(km

)C

ross

line

(km

)

4.54.51.21.2

3.53.5

1.51.5 Inline (km)Inline (km)

Receiver DistributionReceiver DistributionC

ross

line

(km

)C

ross

line

(km

)

4.54.51.21.2

3.53.5

1.51.5 Inline (km)Inline (km)

Inline KM Inline KM ((CPU=1CPU=1)) Inline WMInline WM ((CPU=1/14CPU=1/14))

Offset (km)Offset (km)0.40.4 4.54.5

0.80.8

3.83.8

De

pth

(k

m)

De

pth

(k

m)

Offset (km)Offset (km)0.40.4 4.54.5

Inline KMInline KM ((CPU=1CPU=1)) Inline WMInline WM ((CPU=1/50CPU=1/50))

Offset (km)Offset (km)0.40.4 4.54.5

0.80.8

3.83.8

De

pth

(k

m)

De

pth

(k

m)

Offset (km)Offset (km)0.40.4 4.54.5

(subsample)(subsample)

Crossline KM Crossline KM ((CPU=1CPU=1)) Crossline WMCrossline WM ((CPU=1/14CPU=1/14))

Offset (km)Offset (km)0.30.3 3.53.5

0.80.8

3.33.3

De

pth

(k

m)

De

pth

(k

m)

Offset (km)Offset (km)0.30.3 3.53.5

Crossline KMCrossline KM ((CPU=1CPU=1)) Crossline WMCrossline WM ((CPU=1/50CPU=1/50))(subsample)(subsample)

Offset (km)Offset (km)0.30.3 3.53.5

0.80.8

3.33.3

De

pth

(k

m)

De

pth

(k

m)

Offset (km)Offset (km)0.30.3 3.53.5

Inline: 0~4.6 km, Crossline: 0~3.8Inline: 0~4.6 km, Crossline: 0~3.8

KM (KM (CPU=1CPU=1))

Horizontal Slices (Depth=2.5 km) Horizontal Slices (Depth=2.5 km)

WM (WM (CPU=1/14CPU=1/14)) WM (Sub, WM (Sub, CPU=1/50CPU=1/50))

OutlineOutline

• Problems in 2-D WMProblems in 2-D WM• Objectives of 3-D WM Objectives of 3-D WM • Numerical ResultsNumerical Results ConclusionsConclusions

ConclusionsConclusions

SEG/EAGE Salt DataSEG/EAGE Salt Data• Fewer Migration ArtifactsFewer Migration Artifacts• Better for Complex Salt BoundaryBetter for Complex Salt Boundary• Higher Computational EfficiencyHigher Computational Efficiency

CPUCPU KM: KM: 11 WM: WM: 1/331/33 Subsampled WM: Subsampled WM: 1/1701/170

ConclusionsConclusions

West Texas Field DataWest Texas Field Data• Fewer Migration ArtifactsFewer Migration Artifacts• Similar Image QualitySimilar Image Quality• Higher Computational EfficiencyHigher Computational Efficiency

CPUCPU KM: KM: 11 WM: WM: 1/141/14 Subsampled WM: Subsampled WM: 1/501/50

ConclusionsConclusions

• Trade-off:Trade-off: between Image Quality between Image Quality and CPU Costsand CPU Costs

• Caution:Caution: WM Angle Estimation WM Angle Estimation Sensitive to Recording Geometry Sensitive to Recording Geometry

• More Robust Angle CalculationMore Robust Angle Calculation• Crossing-event CalculationCrossing-event Calculation• 3-D Marine Field Data3-D Marine Field Data• 3-D Iterative Velocity Analysis3-D Iterative Velocity Analysis

Future WorkFuture Work

Real-time Velocity UpdatingReal-time Velocity Updatingfor Target-oriented Migrationfor Target-oriented Migration

Possible ?Possible ?

QuestionQuestion

AcknowledgementsAcknowledgements

I thank UTAM sponsorsI thank UTAM sponsorsfor their financial supportfor their financial support