Waveform Inversion of 4D Seismic Data Using the Ray-born Approximation in the Frequency Domain

Jakobsen, Morten; Keers, Henk; Ruud, Bent Ole; Pšenčı́k, Ivan; Shahraini, A.

doi:10.3997/2214-4609.201401207

Cited by 3 publications

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The T-matrix approach is based on the decomposition of the actual medium into a relatively simple background medium for which the Green's functions are (approximately) known, and a perturbation representing the TR. Since the background medium can be selected arbitrarily and even updated during the inversion process (Oristaglio and Blok, 2012), we think that our method will find many useful applications within timelapse seismics (e.g., Kirchher and Shapiro, 2001;Jakobsen et al, 2010;Jakobsen and Ursin, 2011;Jakobsen, 2012) as well as traditional seismic imaging inversion.…”

Section: Introductionmentioning

confidence: 98%

“…From the viewpoint of the Lipmann-Schwinger integral equation (Oristaglio and Blok, 2012;Jakobsen, 2012), an exact solution requires knowing the total field inside the TR which is impossible. In methods based on the (linearized) Born approximation (e.g., Jakobsen et al, 2010), the total field inside the TR is assumed equal to the incident field. In cases where the (contrast is not sufficiently small) TR is strongly scattering, the assumption breaks down, thus making the simple (single scattering) Born approximation useless for many practical imaging problems (see Malcholm and de Hoop, 2005;Malcholm et al, 2009;Jakobsen, 2012;Oristaglio and Blok, 2012).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Nonlinear seismic waveform inversion using a Born iterative T-matrix method

Jakobsen

Ursin²

2012

SEG Technical Program Expanded Abstracts 2012

View full text Add to dashboard Cite

We present a solution to the nonlinear seismic waveform inversion problem which is based on a T-matrix approach from multiple scattering theory. The solution is presented in the frequency domain, and all frequencies within the seismic bandwidth can in principle be used in the inversion process. The algorithm effectively replaces the nonlinear inverse problem with a sequence of linear inverse problems for which there already exists efficient solution methods (e.g., regularized Born inversion or migration).The T-matrix approach is used to update the field inside the scattering domain after each iteration. This means that internal multiples are not treated as noise, but used to obtain an improved illumunation and imaging of complex targets. In a set of numerical experiments assocated with a 2D seismic model, this Born iterative T-matrix method converged to the true seismic velocity model after a relatively low number of iterations, even in the presence of relatively large contrasts.

show abstract

Section: Introductionmentioning

confidence: 98%