Constrained full-waveform inversion by model reparameterization

Guitton, Antoine; Ayeni, Gboyega; Díaz, Esteban

doi:10.1190/geo2011-0196.1

Cited by 119 publications

(36 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Similar to the methodology of Wang and Tsvankin (2013), we apply image-guided smoothing (Hale, 2009) to the VSP-data gradients to propagate the updates away from the borehole locations. This approach is close to that described by Guitton et al (2012) for isotropic FWI. The gradients are used in the L-BFGS (Nocedal and Wright, 2006) model-updating algorithm.…”

Section: Inversion Algorithmmentioning

confidence: 61%

Acoustic VTI wavefield tomography of P-wave surface and VSP data

Tsvankin

Guitton

et al. 2017

SEG Technical Program Expanded Abstracts 2017

Self Cite

View full text Add to dashboard Cite

CitationLi V, Tsvankin I, Guitton A, Alkhalifah T (2017) Acoustic VTI wavefield tomography of P-wave surface and VSP data. SUMMARYTransversely isotropic (TI) models have become standard in depth imaging and are often used in waveform inversion. Here, we develop a robust wave-equation-based tomographic algorithm for building acoustic VTI (transversely isotropic with a vertical symmetry axis) velocity models from P-wave surface reflection and vertical seismic profiling (VSP) data. Wavefield extrapolation is performed with an integral operator to avoid generating shear-wave artifacts. Focusing energy in extended images produced by reverse-time migration (RTM) makes it possible to update the zero-dip NMO velocity V nmo and the anellipiticity parameter η. To constrain the anisotropy coefficient δ and improve the accuracy in V nmo and η, we employ borehole information by introducing an additional objectivefunction term designed to fit VSP data. Image-guided smoothing is applied to both data-and image-domain gradients to steer the inversion towards geologically plausible solutions.Testing on the VTI Marmousi model shows that the joint inversion of surface and VSP data helps estimate all three relevant medium parameters.

show abstract

Section: Inversion Algorithmmentioning

confidence: 61%

Acoustic VTI wavefield tomography of P-wave surface and VSP data

Tsvankin

Guitton

et al. 2017

SEG Technical Program Expanded Abstracts 2017

Self Cite

View full text Add to dashboard Cite

show abstract

“…The relationship between d and P is dðx; t; x s Þ ¼ Pðx; t; x s Þj x¼x r . Some preconditioning methods can also be applied to the gradient to improve the convergence and to mitigate the nonlinearity of the inverse problem, for example, smoothing and/or filtering the gradient in the wavenumber domain (Nemeth et al, 1997;Guitton et al, 2012;Ma et al, 2012) and using depth-dependent weights for the gradient (Shipp and Singh, 2002;Wang and Rao, 2009). Moreover, the gradient can be normalized by the amplitude of the forward wavefield to approximately account for geometrical divergence (Gauthier et al, 1986) …”

Section: Objective Functionmentioning

confidence: 99%

Viscoacoustic waveform inversion of velocity structures in the time domain

et al. 2014

View full text Add to dashboard Cite

Because of the conversion of elastic energy into heat, seismic waves are attenuated and dispersed as they propagate. The attenuation effects can reduce the resolution of velocity models obtained from waveform inversion or even cause the inversion to produce incorrect results. Using a viscoacoustic model consisting of a single standard linear solid, we discovered a theoretical framework of viscoacoustic waveform inversion in the time domain for velocity estimation. We derived and found the viscoacoustic wave equations for forward modeling and their adjoint to compensate for the attenuation effects in viscoacoustic waveform inversion. The wave equations were numerically solved by high-order finite-difference methods on centered grids to extrapolate seismic wavefields. The finite-difference methods were implemented satisfying stability conditions, which are also presented. Numerical examples proved that the forward viscoacoustic wave equation can simulate attenuative behaviors very well in amplitude attenuation and phase dispersion. We tested acoustic and viscoacoustic waveform inversions with a modified Marmousi model and a 3D field data set from the deep-water Gulf of Mexico for comparison. The tests with the modified Marmousi model illustrated that the seismic attenuation can have large effects on waveform inversion and that choosing the most suitable inversion method was important to obtain the best inversion results for a specific seismic data volume. The tests with the field data set indicated that the inverted velocity models determined from the acoustic and viscoacoustic inversions were helpful to improve images and offset gathers obtained from migration. Compared to the acoustic inversion, viscoacoustic inversion is a realistic approach for real earth materials because the attenuation effects are compensated.

show abstract

“…At further expanse, the application of gradient-based methods allows for updating model m est in several linear iterations (e.g., Pratt, Shin and Hicks 1998). Therefore, the building of model m est is fully integrable with full-waveform inversion technology (e.g., Tarantola 1984;Pratt 1999;Plessix 2006;Virieux and Operto 2009;Zhu et al 2009), which has shown to produce high-resolution models in recent case studies (e.g., Sirgue et al 2010;Vigh et al 2010;Guitton, Ayeni and Díaz 2012). Alternatively, tools in velocity model building based on the input of a human interpreter can be used for creating a model with sharp interfaces.…”

Section: Wavefield Extrapolationmentioning

confidence: 99%

Increasing illumination and sensitivity of reverse‐time migration with internal multiples

Fleury

2013

Geophysical Prospecting

View full text Add to dashboard Cite

A B S T R A C TReverse-time migration is a two-way time-domain finite-frequency technique that accurately handles the propagation of complex scattered waves and produces a bandlimited representation of the subsurface structure that is conventionally assumed to be linear in the contrasts in model parameters. Because of this underlying linear singlescattering assumption, most implementations of this method do not satisfy the energy conservation principle and do not optimally use illumination and model sensitivity of multiply scattered waves. Migrating multiply scattered waves requires preserving the non-linear relation between the image and perturbation of model parameters. I modify the extrapolation of source and receiver wavefields to more accurately handle multiply scattered waves. I extend the concept of the imaging condition in order to map into the subsurface structurally coherent seismic events that correspond to the interaction of both singly and multiply scattered waves. This results in an imaging process referred to here as non-linear reverse-time migration. It includes a strategy that analyses separated contributions of singly and multiply scattered waves to a final non-linear image. The goal is to provide a tool suitable for seismic interpretation and potentially migration velocity analysis that benefits from increased illumination and sensitivity from multiply scattered seismic waves. It is noteworthy that this method can migrate internal multiples, a clear advantage for imaging challenging complex subsurface features, e.g., in salt and basalt environments. The results of synthetic seismic imaging experiments, including a subsalt imaging example, illustrate the technique.

show abstract

Constrained full-waveform inversion by model reparameterization

Cited by 119 publications

References 33 publications

Acoustic VTI wavefield tomography of P-wave surface and VSP data

Acoustic VTI wavefield tomography of P-wave surface and VSP data

Viscoacoustic waveform inversion of velocity structures in the time domain

Increasing illumination and sensitivity of reverse‐time migration with internal multiples

Contact Info

Product

Resources

About