Velocities, Imaging, and Waveform Inversion

The evolution of characterizing the Earth’s subsurface (EET 13)

image of Velocities, Imaging, and Waveform Inversion
  • By Ian F. Jones
  • Format: EPUB
  • Publication Year: 2018
  • Number of Pages: 238
  • Language: English
  • Ebook ISBN: 9789462822535

Velocities, Imaging, and Waveform Inversion - The evolution of characterizing the Earth’s subsurface is part of Ian Jones' EAGE Education Tour and will be a fusion of practical industrial elements, concentrating on the origin and nature of the geological complexities that give rise to imaging problems, as well as a physical (rather than mathematical) understanding of subsurface parameter estimation, and will also look at some possible future directions. The course is designed for: practising geoscientists who desire to better understand the principles and limitations of both current and emerging technologies involved in subsurface parameter estimation and imaging and geoscience students. Following this course, participants should ideally understand how contemporary velocity estimation methods work, and what approximations are involved in obtaining computationally tractable solutions.
In using sound waves to characterise the Earth’s subsurface, we can employ ray theory and/or wave theory, and both migration algorithms and parameter estimation schemes employ one or other of these theoretical descriptions. In this course, we will review the evolution of the industry’s approaches to building earth models via velocity estimation and imaging, outlining the evolution from ray tomography to full waveform inversion, and look towards the emerging possibilities for replacing imaging techniques with direct subsurface parameter inversion methods.
The approach will be mostly non-mathematical, concentrating on an intuitive understanding of the principles, demonstrating them via case histories, and will be divided into the following sections:
- Dealing with the near surface
- The effects of strong vertical velocity contrasts
- The effects of strong lateral velocity contrasts
- Waves versus rays - Model building using ray methods (tomography)
- Model building using wavefield extrapolation methods (FWI)
- Data examples and comparisons
- Future developments
The first three sections outline the nature of the problems we face when building images representing subsurface impedance contrasts, and the next three deal with the technology we deploy to address the problems. In addition, I have included three appendices to outline: the historical development of model building, anisotropy and pre-processing considerations for complex imaging. Several of the individual chapters build on a series of recent tutorial papers which I published in First Break. However, only the key points from these tutorial papers are included, so I refer readers to the original papers for more detail and/or a range of real data examples for each of their topics.
However, due to space and time constraints in the EET format, I have had to omit or limit coverage of various topics, including migration of multiples, Marchenko and inverse scattering series migration, joint migration-inversion, least-squares migration and uncertainty estimation.

Table of Contents

Table of Contents

Chapter 1: Introduction
The need for a geophysical subsurface model
What migration sets out to do
The ‘v’ word
Deriving parameters for migration: 1D versus 3D assumptions
The historical development of velocity model building

Chapter 2: The near surface
Q: ‘when is a push-down not a push-down?
A: ‘when it is a pull-up!Land environments: topography and statics
Migration algorithm statics
Methods for estimating geobody velocity structure
Method 1: measureable velocity error on CMP gathers and use of conventional tomography
Method 2: discernible geobody geometry
Method 3: deeper geometric distortion
Method 4: refraction tomography
Method 5: waveform inversion
Inversion of surface wave data

Chapter 3: Strong vertical velocity change
Modelling study
Observations from acoustic modelling
Simplified modelling to help understand complex arrivals
Converted mode arrivals
Alternative S velocity model
Acoustic and elastic modelling, with and without surface multiples
Comparison with real data
More on converted mode events
Using, rather than removing converted mode events

Chapter 4: Strong lateral velocity change
Physical properties of evaporites
Problems and pitfalls with seismic imaging of salt diapirs
One-way versus two-way depth imaging
The imaging condition in shot migration
RTM imaging artefacts in the vicinity of rapid lateral velocity change
Pitfalls of one-way migration of two-way raypaths
Modelling studies
How complex can salt-related seismic arrivals be?
Refracted energy
Wave mode conversion
Stress and buoyancy effects

Chapter 5: Building models with waves or rays
Resolution scale length
Model update
Ray theory
Generic iterative model building loop for ray-based tomography
Iterative preSDM model update
Tomography cell size variation
Wavefield extrapolation

Chapter 6: The mechanics of ray tomography
What is tomography?
Types and domains of tomography
Travel-time (ray) tomography
Travel-time (ray) tomography in the data domain
Travel-time (ray) tomography in the migrated domain
Some practical aspects of tomographic model updating
Data pre-conditioning
Parametric versus non-parametric picking
Structural constraints identifying layers and faults

Chapter 7: The mechanics of waveform inversion
Using wavefields to build images
The imaging condition for wavefield extrapolation methods
1D convolutional model analogy
Crosstalk imaging artefacts in RTM
Using the rtm imaging condition to perform waveform inversion
Line search
Line search step length
The appearance of the gradient
The components of the gradient
‘One man’s meat is another man’s poison’
The form of the residual
Reconstructed wavefield FWI
Some practical aspects
Finite difference modelling and the acoustic approximation
Introducing density: the quasi-elastic approach
Pre-processing considerations
Frequency selection strategies
Offset selection strategies
Shot decimation
Gradient conditioning
Wavelet estimation and conditioning
QC procedures

Chapter 8: Data comparisons
State of the art with ray tomography
Tomography versus FWI: land and shallow water examples
Shell: land shale and carbonate layering FWI, Oman
Centrica and ION: FWI with strong vertical velocity variation, Norway
Imperial College: variable density FWI, shallow gas leakage
ION and BP, Valhall gas leakage
Shell: shallow gas visco-acoustic FWI, Brunei
Tomography versus FWI: deep water sediment examples
PGS: decomposed gradient update example, GOM
Schlumberger WesternGeco: mobilized shale, GOM
Ophir and ION: gas-charged channels, Equatorial Guinea
Total: gas anomalies, Black Sea
Statoil and ION: silaceous ooze geobodies, Nyq High, Norway
ION: overburden gas accumulation FWI, Anegada, GOM
Woodside energy and CGG: shelf slope imaging, Australia
Chevron and CGG: paleo-canyon distortions, Campos Basin
CGG: refraction and reflection FWI, mobilized shale, GOM
Tomography versus FWI: deep water salt examples
ION and BP: reconstructed wavefield FWI, Messinian salt Nile Delta
BP: Atlantis low frequency FWI sub-salt, GOM
CGG: RFWI sub-salt, GOM
Schlumberger WesternGeco: adjustive FWI sub-salt, GOM

Chapter 9: Conclusions and future development
Tomography versus FWI?
Improving the image
Least-squares migration
Extending FWI
Alternatives to the FWI framework and imaging with multiples
Estimating image uncertainty
What comes next?

Appendix 1: Velocity Model Update Through The Ages
Isotropic model building and ‘depthing’
Model updating for ray-based methods: picking and inverting
Evolution of non-tomographic techniques
Map migration
Coherency inversion
The Deregowski loop
Wavefield extrapolation and focusing analysis
CRP gather scanning and image scanning techniques
Evolution of tomographic techniques
CFP analysis
CRS picking and multifocusing

Appendix 2: Anisotropy
Manifestation in seismic data
The origin of stratigraphically induced anisotropy
Orthorhombic TTI: adding orthogonal fracture sets to the TI model
Stress-induced anisotropy
Wide-azimuth and multi-azimuth data
Anisotropic model building
Anisotropic preSDM in the absence of well control

Appendix 3: Pre-processing considerations for RTM
Modelling results
Effects on RTM migration



This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error