Event-consistent smoothing in generalized Introduction Conventional - - PowerPoint PPT Presentation

▶
event consistent smoothing in generalized
SMART_READER_LITE
LIVE PREVIEW

Event-consistent smoothing in generalized Introduction Conventional - - PowerPoint PPT Presentation

Nov 15, 2022 545 likes •1.49k views

74 th Annual Meeting SEG, Denver 2004 Mann & Duveneck Event-consistent smoothing in generalized Introduction Conventional CRS stack high-density velocity analysis Why smoothing? Pulse stretch Smoothing algorithm E. Duveneck 1 J. Mann

slide-1
SLIDE 1

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Event-consistent smoothing in generalized high-density velocity analysis

  • J. Mann
  • E. Duveneck1

1now: SINTEF Petroleum Research, Trondheim, Norway

Wave Inversion Technology (WIT) Consortium Geophysical Institute, University of Karlsruhe (TH) October 12, 2004

slide-2
SLIDE 2

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Overview

Introduction Smoothing algorithm Data examples Conclusions Acknowledgments

slide-3
SLIDE 3

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Introduction

Conventional stacking velocity analysis:

◮ (semi-)interactive, interpretative velocity picking ◮ coarse picks on selected key events, only

☞ human interaction required ☞ low temporal and spatial resolution ☞ pulse stretch deteriorates stack result Thus desirable:

◮ automated approach ◮ more appropriate parameterization ◮ maximum resolution

slide-4
SLIDE 4

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Introduction

Conventional stacking velocity analysis:

◮ (semi-)interactive, interpretative velocity picking ◮ coarse picks on selected key events, only

☞ human interaction required ☞ low temporal and spatial resolution ☞ pulse stretch deteriorates stack result Thus desirable:

◮ automated approach ◮ more appropriate parameterization ◮ maximum resolution

slide-5
SLIDE 5

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Introduction

Conventional stacking velocity analysis:

◮ (semi-)interactive, interpretative velocity picking ◮ coarse picks on selected key events, only

☞ human interaction required ☞ low temporal and spatial resolution ☞ pulse stretch deteriorates stack result Thus desirable:

◮ automated approach ◮ more appropriate parameterization ◮ maximum resolution

slide-6
SLIDE 6

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Introduction

Conventional stacking velocity analysis:

◮ (semi-)interactive, interpretative velocity picking ◮ coarse picks on selected key events, only

☞ human interaction required ☞ low temporal and spatial resolution ☞ pulse stretch deteriorates stack result Thus desirable:

◮ automated approach ◮ more appropriate parameterization ◮ maximum resolution

slide-7
SLIDE 7

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Introduction

Conventional stacking velocity analysis:

◮ (semi-)interactive, interpretative velocity picking ◮ coarse picks on selected key events, only

☞ human interaction required ☞ low temporal and spatial resolution ☞ pulse stretch deteriorates stack result Thus desirable:

◮ automated approach ◮ more appropriate parameterization ◮ maximum resolution

slide-8
SLIDE 8

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Introduction

Conventional stacking velocity analysis:

◮ (semi-)interactive, interpretative velocity picking ◮ coarse picks on selected key events, only

☞ human interaction required ☞ low temporal and spatial resolution ☞ pulse stretch deteriorates stack result Thus desirable:

◮ automated approach ◮ more appropriate parameterization ◮ maximum resolution

slide-9
SLIDE 9

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Introduction

Conventional stacking velocity analysis:

◮ (semi-)interactive, interpretative velocity picking ◮ coarse picks on selected key events, only

☞ human interaction required ☞ low temporal and spatial resolution ☞ pulse stretch deteriorates stack result Thus desirable:

◮ automated approach ◮ more appropriate parameterization ◮ maximum resolution

slide-10
SLIDE 10

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Introduction

Conventional stacking velocity analysis:

◮ (semi-)interactive, interpretative velocity picking ◮ coarse picks on selected key events, only

☞ human interaction required ☞ low temporal and spatial resolution ☞ pulse stretch deteriorates stack result Thus desirable:

◮ automated approach ◮ more appropriate parameterization ◮ maximum resolution

slide-11
SLIDE 11

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Introduction

Conventional stacking velocity analysis:

◮ (semi-)interactive, interpretative velocity picking ◮ coarse picks on selected key events, only

☞ human interaction required ☞ low temporal and spatial resolution ☞ pulse stretch deteriorates stack result Thus desirable:

◮ automated approach ◮ more appropriate parameterization ◮ maximum resolution

slide-12
SLIDE 12

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Introduction

Conventional stacking velocity analysis:

◮ (semi-)interactive, interpretative velocity picking ◮ coarse picks on selected key events, only

☞ human interaction required ☞ low temporal and spatial resolution ☞ pulse stretch deteriorates stack result Thus desirable:

◮ automated approach ◮ more appropriate parameterization ◮ maximum resolution

slide-13
SLIDE 13

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Common-Reflection-Surface (CRS) stack

Generalization of conventional approach:

◮ second-order approximation of traveltime ◮ fully automated coherence-based application ◮ high-density analysis ◮ spatial stacking operator

☞ much more prestack traces used ☞ enhanced signal/noise ratio

◮ additional stacking parameters related to 1. and 2.

traveltime derivatives ☞ geometrical interpretation

slide-14
SLIDE 14

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Common-Reflection-Surface (CRS) stack

Generalization of conventional approach:

◮ second-order approximation of traveltime ◮ fully automated coherence-based application ◮ high-density analysis ◮ spatial stacking operator

☞ much more prestack traces used ☞ enhanced signal/noise ratio

◮ additional stacking parameters related to 1. and 2.

traveltime derivatives ☞ geometrical interpretation

slide-15
SLIDE 15

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Common-Reflection-Surface (CRS) stack

Generalization of conventional approach:

◮ second-order approximation of traveltime ◮ fully automated coherence-based application ◮ high-density analysis ◮ spatial stacking operator

☞ much more prestack traces used ☞ enhanced signal/noise ratio

◮ additional stacking parameters related to 1. and 2.

traveltime derivatives ☞ geometrical interpretation

slide-16
SLIDE 16

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Common-Reflection-Surface (CRS) stack

Generalization of conventional approach:

◮ second-order approximation of traveltime ◮ fully automated coherence-based application ◮ high-density analysis ◮ spatial stacking operator

☞ much more prestack traces used ☞ enhanced signal/noise ratio

◮ additional stacking parameters related to 1. and 2.

traveltime derivatives ☞ geometrical interpretation

slide-17
SLIDE 17

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Common-Reflection-Surface (CRS) stack

Generalization of conventional approach:

◮ second-order approximation of traveltime ◮ fully automated coherence-based application ◮ high-density analysis ◮ spatial stacking operator

☞ much more prestack traces used ☞ enhanced signal/noise ratio

◮ additional stacking parameters related to 1. and 2.

traveltime derivatives ☞ geometrical interpretation

slide-18
SLIDE 18

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Common-Reflection-Surface (CRS) stack

Generalization of conventional approach:

◮ second-order approximation of traveltime ◮ fully automated coherence-based application ◮ high-density analysis ◮ spatial stacking operator

☞ much more prestack traces used ☞ enhanced signal/noise ratio

◮ additional stacking parameters related to 1. and 2.

traveltime derivatives ☞ geometrical interpretation

slide-19
SLIDE 19

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Common-Reflection-Surface (CRS) stack

Generalization of conventional approach:

◮ second-order approximation of traveltime ◮ fully automated coherence-based application ◮ high-density analysis ◮ spatial stacking operator

☞ much more prestack traces used ☞ enhanced signal/noise ratio

◮ additional stacking parameters related to 1. and 2.

traveltime derivatives ☞ geometrical interpretation

slide-20
SLIDE 20

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Common-Reflection-Surface (CRS) stack

Generalization of conventional approach:

◮ second-order approximation of traveltime ◮ fully automated coherence-based application ◮ high-density analysis ◮ spatial stacking operator

☞ much more prestack traces used ☞ enhanced signal/noise ratio

◮ additional stacking parameters related to 1. and 2.

traveltime derivatives ☞ geometrical interpretation

slide-21
SLIDE 21

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Common-Reflection-Surface stack

Geometrical interpretation of stacking parameters:

x0 x0 RNIP

N

R

α α

NIP NIP

Emergence direction and curvatures of hypothetical wavefronts:

◮ exploding point source ☞ normal-incidence-point

(NIP) wave

◮ exploding reflector ☞ normal (N) wave

slide-22
SLIDE 22

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Common-Reflection-Surface stack

Geometrical interpretation of stacking parameters:

x0 x0 RNIP

N

R

α α

NIP NIP

Emergence direction and curvatures of hypothetical wavefronts:

◮ exploding point source ☞ normal-incidence-point

(NIP) wave

◮ exploding reflector ☞ normal (N) wave

slide-23
SLIDE 23

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Common-Reflection-Surface stack

Geometrical interpretation of stacking parameters:

x0 x0 RNIP

N

R

α α

NIP NIP

Emergence direction and curvatures of hypothetical wavefronts:

◮ exploding point source ☞ normal-incidence-point

(NIP) wave

◮ exploding reflector ☞ normal (N) wave

slide-24
SLIDE 24

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Common-Reflection-Surface stack

Geometrical interpretation of stacking parameters:

x0 x0 RNIP

N

R

α α

NIP NIP

Emergence direction and curvatures of hypothetical wavefronts:

◮ exploding point source ☞ normal-incidence-point

(NIP) wave

◮ exploding reflector ☞ normal (N) wave

slide-25
SLIDE 25

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

High-density analysis vs. smoothing

Stacking parameters are subject to

◮ fluctuations due to noise ◮ outliers due to failures to detect the relevant

coherence maximum Stacking parameters represent integral properties of the subsurface ➥ smooth variation along reflection events ➥ event-consistent smoothing along reflection events is justified!

slide-26
SLIDE 26

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

High-density analysis vs. smoothing

Stacking parameters are subject to

◮ fluctuations due to noise ◮ outliers due to failures to detect the relevant

coherence maximum Stacking parameters represent integral properties of the subsurface ➥ smooth variation along reflection events ➥ event-consistent smoothing along reflection events is justified!

slide-27
SLIDE 27

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

High-density analysis vs. smoothing

Stacking parameters are subject to

◮ fluctuations due to noise ◮ outliers due to failures to detect the relevant

coherence maximum Stacking parameters represent integral properties of the subsurface ➥ smooth variation along reflection events ➥ event-consistent smoothing along reflection events is justified!

slide-28
SLIDE 28

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

High-density analysis vs. smoothing

Stacking parameters are subject to

◮ fluctuations due to noise ◮ outliers due to failures to detect the relevant

coherence maximum Stacking parameters represent integral properties of the subsurface ➥ smooth variation along reflection events ➥ event-consistent smoothing along reflection events is justified!

slide-29
SLIDE 29

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

High-density analysis vs. smoothing

Stacking parameters are subject to

◮ fluctuations due to noise ◮ outliers due to failures to detect the relevant

coherence maximum Stacking parameters represent integral properties of the subsurface ➥ smooth variation along reflection events ➥ event-consistent smoothing along reflection events is justified!

slide-30
SLIDE 30

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

High-density analysis vs. smoothing

Stacking parameters are subject to

◮ fluctuations due to noise ◮ outliers due to failures to detect the relevant

coherence maximum Stacking parameters represent integral properties of the subsurface ➥ smooth variation along reflection events ➥ event-consistent smoothing along reflection events is justified!

slide-31
SLIDE 31

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

High-density analysis vs. smoothing

Bandwidth is limited. What happens along the wavelet?

◮ high-density stacking velocity

◮ systematic variation along wavelet ◮ smoothing reintroduces pulse stretch phenomenon

◮ CRS stacking parameters

◮ virtually constant along wavelet ◮ smoothing also allowed along wavelet without pulse

stretch

slide-32
SLIDE 32

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

High-density analysis vs. smoothing

Bandwidth is limited. What happens along the wavelet?

◮ high-density stacking velocity

◮ systematic variation along wavelet ◮ smoothing reintroduces pulse stretch phenomenon

◮ CRS stacking parameters

◮ virtually constant along wavelet ◮ smoothing also allowed along wavelet without pulse

stretch

slide-33
SLIDE 33

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

High-density analysis vs. smoothing

Bandwidth is limited. What happens along the wavelet?

◮ high-density stacking velocity

◮ systematic variation along wavelet ◮ smoothing reintroduces pulse stretch phenomenon

◮ CRS stacking parameters

◮ virtually constant along wavelet ◮ smoothing also allowed along wavelet without pulse

stretch

slide-34
SLIDE 34

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

High-density analysis vs. smoothing

Bandwidth is limited. What happens along the wavelet?

◮ high-density stacking velocity

◮ systematic variation along wavelet ◮ smoothing reintroduces pulse stretch phenomenon

◮ CRS stacking parameters

◮ virtually constant along wavelet ◮ smoothing also allowed along wavelet without pulse

stretch

slide-35
SLIDE 35

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

High-density analysis vs. smoothing

Bandwidth is limited. What happens along the wavelet?

◮ high-density stacking velocity

◮ systematic variation along wavelet ◮ smoothing reintroduces pulse stretch phenomenon

◮ CRS stacking parameters

◮ virtually constant along wavelet ◮ smoothing also allowed along wavelet without pulse

stretch

slide-36
SLIDE 36

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

High-density analysis vs. smoothing

Bandwidth is limited. What happens along the wavelet?

◮ high-density stacking velocity

◮ systematic variation along wavelet ◮ smoothing reintroduces pulse stretch phenomenon

◮ CRS stacking parameters

◮ virtually constant along wavelet ◮ smoothing also allowed along wavelet without pulse

stretch

slide-37
SLIDE 37

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

High-density analysis vs. smoothing

Bandwidth is limited. What happens along the wavelet?

◮ high-density stacking velocity

◮ systematic variation along wavelet ◮ smoothing reintroduces pulse stretch phenomenon

◮ CRS stacking parameters

◮ virtually constant along wavelet ◮ smoothing also allowed along wavelet without pulse

stretch

slide-38
SLIDE 38

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Pulse stretch phenomenon

Smooth model: stacking velocity vs. CRS parameters

1 1.1 1.2 1.3 1.4 1.5 100 200 300 400 500 600 700

Stretch factor Half−Offset [m] iso−phase trajecories constant velocity velocity gradient constant ZO curvature From: Mann and Höcht, 2003, Pulse stretch effects in the context of data-driven imaging methods, 65th Conf., Eur. Assn. Geosci. Eng.

slide-39
SLIDE 39

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Requirements:

◮ smoothing along reflection events justified

✔

◮ smoothing along wavelet justified

✔

◮ remaining task: ensure event consistence

CRS stack provides:

◮ local shape of zero-offset reflection event (α, RN) ◮ approximation of projected Fresnel zone ◮ coherence values as measure of reliability

☞ this allows a simple and efficient smoothing algorithm

slide-40
SLIDE 40

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Requirements:

◮ smoothing along reflection events justified

✔

◮ smoothing along wavelet justified

✔

◮ remaining task: ensure event consistence

CRS stack provides:

◮ local shape of zero-offset reflection event (α, RN) ◮ approximation of projected Fresnel zone ◮ coherence values as measure of reliability

☞ this allows a simple and efficient smoothing algorithm

slide-41
SLIDE 41

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Requirements:

◮ smoothing along reflection events justified

✔

◮ smoothing along wavelet justified

✔

◮ remaining task: ensure event consistence

CRS stack provides:

◮ local shape of zero-offset reflection event (α, RN) ◮ approximation of projected Fresnel zone ◮ coherence values as measure of reliability

☞ this allows a simple and efficient smoothing algorithm

slide-42
SLIDE 42

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Requirements:

◮ smoothing along reflection events justified

✔

◮ smoothing along wavelet justified

✔

◮ remaining task: ensure event consistence

CRS stack provides:

◮ local shape of zero-offset reflection event (α, RN) ◮ approximation of projected Fresnel zone ◮ coherence values as measure of reliability

☞ this allows a simple and efficient smoothing algorithm

slide-43
SLIDE 43

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Requirements:

◮ smoothing along reflection events justified

✔

◮ smoothing along wavelet justified

✔

◮ remaining task: ensure event consistence

CRS stack provides:

◮ local shape of zero-offset reflection event (α, RN) ◮ approximation of projected Fresnel zone ◮ coherence values as measure of reliability

☞ this allows a simple and efficient smoothing algorithm

slide-44
SLIDE 44

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Requirements:

◮ smoothing along reflection events justified

✔

◮ smoothing along wavelet justified

✔

◮ remaining task: ensure event consistence

CRS stack provides:

◮ local shape of zero-offset reflection event (α, RN) ◮ approximation of projected Fresnel zone ◮ coherence values as measure of reliability

☞ this allows a simple and efficient smoothing algorithm

slide-45
SLIDE 45

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Requirements:

◮ smoothing along reflection events justified

✔

◮ smoothing along wavelet justified

✔

◮ remaining task: ensure event consistence

CRS stack provides:

◮ local shape of zero-offset reflection event (α, RN) ◮ approximation of projected Fresnel zone ◮ coherence values as measure of reliability

☞ this allows a simple and efficient smoothing algorithm

slide-46
SLIDE 46

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Requirements:

◮ smoothing along reflection events justified

✔

◮ smoothing along wavelet justified

✔

◮ remaining task: ensure event consistence

CRS stack provides:

◮ local shape of zero-offset reflection event (α, RN) ◮ approximation of projected Fresnel zone ◮ coherence values as measure of reliability

☞ this allows a simple and efficient smoothing algorithm

slide-47
SLIDE 47

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Requirements:

◮ smoothing along reflection events justified

✔

◮ smoothing along wavelet justified

✔

◮ remaining task: ensure event consistence

CRS stack provides:

◮ local shape of zero-offset reflection event (α, RN) ◮ approximation of projected Fresnel zone ◮ coherence values as measure of reliability

☞ this allows a simple and efficient smoothing algorithm

slide-48
SLIDE 48

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Requirements:

◮ smoothing along reflection events justified

✔

◮ smoothing along wavelet justified

✔

◮ remaining task: ensure event consistence

CRS stack provides:

◮ local shape of zero-offset reflection event (α, RN) ◮ approximation of projected Fresnel zone ◮ coherence values as measure of reliability

☞ this allows a simple and efficient smoothing algorithm

slide-49
SLIDE 49

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Requirements:

◮ smoothing along reflection events justified

✔

◮ smoothing along wavelet justified

✔

◮ remaining task: ensure event consistence

CRS stack provides:

◮ local shape of zero-offset reflection event (α, RN) ◮ approximation of projected Fresnel zone ◮ coherence values as measure of reliability

☞ this allows a simple and efficient smoothing algorithm

slide-50
SLIDE 50

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

For each zero-offset sample and each CRS parameter

◮ align smoothing window along reflection event using

emergence angle α (optionally also RN)

◮ reject samples below given coherence threshold ☞

use only reliable attributes

◮ reject samples with dip difference beyond threshold

☞ avoid mixing of intersecting events

◮ apply combined filter:

◮ median filter ☞ remove outliers ◮ averaging ☞ remove fluctuations

◮ assign result to zero-offset sample

slide-51
SLIDE 51

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

For each zero-offset sample and each CRS parameter

◮ align smoothing window along reflection event using

emergence angle α (optionally also RN)

◮ reject samples below given coherence threshold ☞

use only reliable attributes

◮ reject samples with dip difference beyond threshold

☞ avoid mixing of intersecting events

◮ apply combined filter:

◮ median filter ☞ remove outliers ◮ averaging ☞ remove fluctuations

◮ assign result to zero-offset sample

slide-52
SLIDE 52

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

For each zero-offset sample and each CRS parameter

◮ align smoothing window along reflection event using

emergence angle α (optionally also RN)

◮ reject samples below given coherence threshold ☞

use only reliable attributes

◮ reject samples with dip difference beyond threshold

☞ avoid mixing of intersecting events

◮ apply combined filter:

◮ median filter ☞ remove outliers ◮ averaging ☞ remove fluctuations

◮ assign result to zero-offset sample

slide-53
SLIDE 53

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

For each zero-offset sample and each CRS parameter

◮ align smoothing window along reflection event using

emergence angle α (optionally also RN)

◮ reject samples below given coherence threshold ☞

use only reliable attributes

◮ reject samples with dip difference beyond threshold

☞ avoid mixing of intersecting events

◮ apply combined filter:

◮ median filter ☞ remove outliers ◮ averaging ☞ remove fluctuations

◮ assign result to zero-offset sample

slide-54
SLIDE 54

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

For each zero-offset sample and each CRS parameter

◮ align smoothing window along reflection event using

emergence angle α (optionally also RN)

◮ reject samples below given coherence threshold ☞

use only reliable attributes

◮ reject samples with dip difference beyond threshold

☞ avoid mixing of intersecting events

◮ apply combined filter:

◮ median filter ☞ remove outliers ◮ averaging ☞ remove fluctuations

◮ assign result to zero-offset sample

slide-55
SLIDE 55

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

For each zero-offset sample and each CRS parameter

◮ align smoothing window along reflection event using

emergence angle α (optionally also RN)

◮ reject samples below given coherence threshold ☞

use only reliable attributes

◮ reject samples with dip difference beyond threshold

☞ avoid mixing of intersecting events

◮ apply combined filter:

◮ median filter ☞ remove outliers ◮ averaging ☞ remove fluctuations

◮ assign result to zero-offset sample

slide-56
SLIDE 56

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

For each zero-offset sample and each CRS parameter

◮ align smoothing window along reflection event using

emergence angle α (optionally also RN)

◮ reject samples below given coherence threshold ☞

use only reliable attributes

◮ reject samples with dip difference beyond threshold

☞ avoid mixing of intersecting events

◮ apply combined filter:

◮ median filter ☞ remove outliers ◮ averaging ☞ remove fluctuations

◮ assign result to zero-offset sample

slide-57
SLIDE 57

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

For each zero-offset sample and each CRS parameter

◮ align smoothing window along reflection event using

emergence angle α (optionally also RN)

◮ reject samples below given coherence threshold ☞

use only reliable attributes

◮ reject samples with dip difference beyond threshold

☞ avoid mixing of intersecting events

◮ apply combined filter:

◮ median filter ☞ remove outliers ◮ averaging ☞ remove fluctuations

◮ assign result to zero-offset sample

slide-58
SLIDE 58

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Size of smoothing window:

◮ as small as possible, as large as required ◮ temporal extension ≤ wavelet length ◮ lateral extension ≪ projected Fresnel zone, either

fixed or a fraction of approximate Fresnel zone given by CRS parameters Smoothing in the 3D case:

◮ smoothing window is a small volume ◮ same selection criteria as in 2D ◮ combined filter has to be generalized for curvature

matrices and slowness vectors ☞ current research

slide-59
SLIDE 59

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Size of smoothing window:

◮ as small as possible, as large as required ◮ temporal extension ≤ wavelet length ◮ lateral extension ≪ projected Fresnel zone, either

fixed or a fraction of approximate Fresnel zone given by CRS parameters Smoothing in the 3D case:

◮ smoothing window is a small volume ◮ same selection criteria as in 2D ◮ combined filter has to be generalized for curvature

matrices and slowness vectors ☞ current research

slide-60
SLIDE 60

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Size of smoothing window:

◮ as small as possible, as large as required ◮ temporal extension ≤ wavelet length ◮ lateral extension ≪ projected Fresnel zone, either

fixed or a fraction of approximate Fresnel zone given by CRS parameters Smoothing in the 3D case:

◮ smoothing window is a small volume ◮ same selection criteria as in 2D ◮ combined filter has to be generalized for curvature

matrices and slowness vectors ☞ current research

slide-61
SLIDE 61

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Size of smoothing window:

◮ as small as possible, as large as required ◮ temporal extension ≤ wavelet length ◮ lateral extension ≪ projected Fresnel zone, either

fixed or a fraction of approximate Fresnel zone given by CRS parameters Smoothing in the 3D case:

◮ smoothing window is a small volume ◮ same selection criteria as in 2D ◮ combined filter has to be generalized for curvature

matrices and slowness vectors ☞ current research

slide-62
SLIDE 62

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Size of smoothing window:

◮ as small as possible, as large as required ◮ temporal extension ≤ wavelet length ◮ lateral extension ≪ projected Fresnel zone, either

fixed or a fraction of approximate Fresnel zone given by CRS parameters Smoothing in the 3D case:

◮ smoothing window is a small volume ◮ same selection criteria as in 2D ◮ combined filter has to be generalized for curvature

matrices and slowness vectors ☞ current research

slide-63
SLIDE 63

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Size of smoothing window:

◮ as small as possible, as large as required ◮ temporal extension ≤ wavelet length ◮ lateral extension ≪ projected Fresnel zone, either

fixed or a fraction of approximate Fresnel zone given by CRS parameters Smoothing in the 3D case:

◮ smoothing window is a small volume ◮ same selection criteria as in 2D ◮ combined filter has to be generalized for curvature

matrices and slowness vectors ☞ current research

slide-64
SLIDE 64

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Size of smoothing window:

◮ as small as possible, as large as required ◮ temporal extension ≤ wavelet length ◮ lateral extension ≪ projected Fresnel zone, either

fixed or a fraction of approximate Fresnel zone given by CRS parameters Smoothing in the 3D case:

◮ smoothing window is a small volume ◮ same selection criteria as in 2D ◮ combined filter has to be generalized for curvature

matrices and slowness vectors ☞ current research

slide-65
SLIDE 65

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Smoothing algorithm

Size of smoothing window:

◮ as small as possible, as large as required ◮ temporal extension ≤ wavelet length ◮ lateral extension ≪ projected Fresnel zone, either

fixed or a fraction of approximate Fresnel zone given by CRS parameters Smoothing in the 3D case:

◮ smoothing window is a small volume ◮ same selection criteria as in 2D ◮ combined filter has to be generalized for curvature

matrices and slowness vectors ☞ current research

slide-66
SLIDE 66

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Schematic example

1.0 1.5 2.0 2.5 3.0 3.5 Time [s] 2 4 6 8 10 Location [km]

CRS stack section

slide-67
SLIDE 67

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Schematic example

1.0 1.5 2.0 2.5 3.0 3.5 Time [s] 2 4 6 8 10 Location [km]

CRS stack section

slide-68
SLIDE 68

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Schematic example

Coherence

0.05 0.10

Emergence angle [°]

  • 40
  • 20

20 40

NIP wave radius [km]

2.0 2.2 2.4 2.6 2.8

slide-69
SLIDE 69

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Schematic example

Coherence

0.05 0.10

Emergence angle [°]

  • 40
  • 20

20 40

NIP wave radius [km]

2.0 2.2 2.4 2.6 2.8 0.05 0.10

  • 40
  • 20

20 40 2.0 2.2 2.4 2.6 2.8

Coherence-based mask applied

(for visualization, only)

slide-70
SLIDE 70

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Schematic example

Coherence

0.05 0.10

Emergence angle [°]

  • 40
  • 20

20 40

NIP wave radius [km]

2.0 2.2 2.4 2.6 2.8 0.05 0.10

  • 40
  • 20

20 40 2.0 2.2 2.4 2.6 2.8

Smoothing window aligned with reflection event

slide-71
SLIDE 71

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Schematic example

Coherence

0.05 0.10

Emergence angle [°]

  • 40
  • 20

20 40

NIP wave radius [km]

2.0 2.2 2.4 2.6 2.8 0.05 0.10

  • 40
  • 20

20 40 2.0 2.2 2.4 2.6 2.8

Select all samples in window

slide-72
SLIDE 72

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Schematic example

Coherence

0.05 0.10

Emergence angle [°]

  • 40
  • 20

20 40

NIP wave radius [km]

2.0 2.2 2.4 2.6 2.8 0.05 0.10

  • 40
  • 20

20 40 2.0 2.2 2.4 2.6 2.8

Apply coherence threshold and dip difference threshold

slide-73
SLIDE 73

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Schematic example

Coherence

0.05 0.10

Emergence angle [°]

  • 40
  • 20

20 40

NIP wave radius [km]

2.0 2.2 2.4 2.6 2.8 0.05 0.10

  • 40
  • 20

20 40 2.0 2.2 2.4 2.6 2.8

Smoothing:

◮ Sort remaining samples by magnitude

slide-74
SLIDE 74

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Schematic example

Coherence

0.05 0.10

Emergence angle [°]

  • 40
  • 20

20 40

NIP wave radius [km]

2.0 2.2 2.4 2.6 2.8 0.05 0.10

  • 40
  • 20

20 40 2.0 2.2 2.4 2.6 2.8

Smoothing:

◮ Sort remaining samples by magnitude ◮ Average given fraction of samples around median

slide-75
SLIDE 75

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Schematic example

Coherence

0.05 0.10

Emergence angle [°]

  • 40
  • 20

20 40

NIP wave radius [km]

2.0 2.2 2.4 2.6 2.8 0.05 0.10

  • 40
  • 20

20 40 2.0 2.2 2.4 2.6 2.8

Smoothing:

◮ Sort remaining samples by magnitude ◮ Average given fraction of samples around median ◮ Assign result to considered ZO location

slide-76
SLIDE 76

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Schematic example

Coherence

0.05 0.10

Emergence angle [°]

  • 40
  • 20

20 40

NIP wave radius [km]

2.0 2.2 2.4 2.6 2.8 0.05 0.10

  • 40
  • 20

20 40 2.0 2.2 2.4 2.6 2.8

Repeated for all location ☞ smoothed attribute sections

0.05 0.10

  • 40
  • 20

20 40 2.0 2.2 2.4 2.6 2.8

slide-77
SLIDE 77

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Real data example

CRS parameter sections (detail)

1.0 1.5 2.0 2.5

  • 15
  • 10
  • 5

5 10 15 1.0 1.5 2.0 2.5 2000 4000 6000 8000

Emergence angle [◦] NIP wave radius [m] Original parameters as obtained by CRS stack (no coherence mask applied)

slide-78
SLIDE 78

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Real data example

CRS parameter sections (detail)

1.0 1.5 2.0 2.5

  • 15
  • 10
  • 5

5 10 15 1.0 1.5 2.0 2.5 2000 4000 6000 8000

Emergence angle [◦] NIP wave radius [m] Original parameters as obtained by CRS stack (coherence mask applied for display, only)

slide-79
SLIDE 79

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Real data example

CRS parameter sections (detail)

1.0 1.5 2.0 2.5

  • 15
  • 10
  • 5

5 10 15 1.0 1.5 2.0 2.5 2000 4000 6000 8000

Emergence angle [◦] NIP wave radius [m] Smoothed parameters (coherence mask applied for display, only)

slide-80
SLIDE 80

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Real data examples

CRS stack sections (detail I) Stack with original vs. stack with smoothed parameters

slide-81
SLIDE 81

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Real data examples

CRS stack sections (detail II) Stack with original vs. stack with smoothed parameters

slide-82
SLIDE 82

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Conclusions

Smoothing algorithm:

◮ event-consistent smoothing ◮ based on CRS stacking parameters and coherence ◮ removes outliers ◮ removes fluctuations ◮ preserves kinematic properties of reflection events ◮ avoids mixing of intersecting events

➥ improved quality of stacked section ➥ more physical CRS stacking parameter sections for various applications like macromodel determination etc.

slide-83
SLIDE 83

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Conclusions

Smoothing algorithm:

◮ event-consistent smoothing ◮ based on CRS stacking parameters and coherence ◮ removes outliers ◮ removes fluctuations ◮ preserves kinematic properties of reflection events ◮ avoids mixing of intersecting events

➥ improved quality of stacked section ➥ more physical CRS stacking parameter sections for various applications like macromodel determination etc.

slide-84
SLIDE 84

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Conclusions

Smoothing algorithm:

◮ event-consistent smoothing ◮ based on CRS stacking parameters and coherence ◮ removes outliers ◮ removes fluctuations ◮ preserves kinematic properties of reflection events ◮ avoids mixing of intersecting events

➥ improved quality of stacked section ➥ more physical CRS stacking parameter sections for various applications like macromodel determination etc.

slide-85
SLIDE 85

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Conclusions

Smoothing algorithm:

◮ event-consistent smoothing ◮ based on CRS stacking parameters and coherence ◮ removes outliers ◮ removes fluctuations ◮ preserves kinematic properties of reflection events ◮ avoids mixing of intersecting events

➥ improved quality of stacked section ➥ more physical CRS stacking parameter sections for various applications like macromodel determination etc.

slide-86
SLIDE 86

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Conclusions

Smoothing algorithm:

◮ event-consistent smoothing ◮ based on CRS stacking parameters and coherence ◮ removes outliers ◮ removes fluctuations ◮ preserves kinematic properties of reflection events ◮ avoids mixing of intersecting events

➥ improved quality of stacked section ➥ more physical CRS stacking parameter sections for various applications like macromodel determination etc.

slide-87
SLIDE 87

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Conclusions

Smoothing algorithm:

◮ event-consistent smoothing ◮ based on CRS stacking parameters and coherence ◮ removes outliers ◮ removes fluctuations ◮ preserves kinematic properties of reflection events ◮ avoids mixing of intersecting events

➥ improved quality of stacked section ➥ more physical CRS stacking parameter sections for various applications like macromodel determination etc.

slide-88
SLIDE 88

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Conclusions

Smoothing algorithm:

◮ event-consistent smoothing ◮ based on CRS stacking parameters and coherence ◮ removes outliers ◮ removes fluctuations ◮ preserves kinematic properties of reflection events ◮ avoids mixing of intersecting events

➥ improved quality of stacked section ➥ more physical CRS stacking parameter sections for various applications like macromodel determination etc.

slide-89
SLIDE 89

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Conclusions

Smoothing algorithm:

◮ event-consistent smoothing ◮ based on CRS stacking parameters and coherence ◮ removes outliers ◮ removes fluctuations ◮ preserves kinematic properties of reflection events ◮ avoids mixing of intersecting events

➥ improved quality of stacked section ➥ more physical CRS stacking parameter sections for various applications like macromodel determination etc.

slide-90
SLIDE 90

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Acknowledgments

This work was kindly supported by the sponsors of the Wave Inversion Technology (WIT) Consortium, Karlsruhe, Germany

slide-91
SLIDE 91

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T

Related presentations

Session SP 4, Thursday morning: SP 4.4 A seismic reflection imaging workflow based on the common-reflection-surface (CRS) stack: theoretical background and case study SP 4.5 CRS imaging and tomography versus PreSDM: a case history in overthrust geology SP 4.6 CRS stack and redatuming for rugged surface topography: a synthetic data example SP 4.8 3D focusing operator estimation using sparse data

slide-92
SLIDE 92

74th Annual Meeting SEG, Denver 2004 Mann & Duveneck Introduction Conventional CRS stack Why smoothing? Pulse stretch Smoothing algorithm Requirements The algorithm Schematic example Data examples Parameters Stack sections Conclusions Acknowledgments Related talks

W I T