Least-squares reverse time migration in the presence of velocity errors

Yang, Jizhong; Li, Yunyue Elita; Cheng, Arthur; Liu, Yuzhu; Dong, Liangguo

doi:10.1190/geo2019-0005.1

Cited by 24 publications

(4 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, if we have access to ∆d e , then each desired coefficient of ∆m can be computed independently by cross-correlation of the associated forward-propagated source-side wavefield with the backward-propagated receiver-side wavefield. This memory efficiency is a main advantage of our algorithm over the existing algorithms based on model extension that require to keep in memory the model for all of the desired range of subsurface offsets [4,5,7,20,24,31]. We compute the desired search direction δm by averaging the extended coefficients over a ball around each physical point.…”

Section: Computing δM From ∆Mmentioning

confidence: 99%

See 1 more Smart Citation

An extended Gauss-Newton method for full waveform inversion

Gholami¹

2023

Preprint

View full text Add to dashboard Cite

Full Waveform Inversion (FWI) is a large-scale nonlinear ill-posed problem for which implementation of the Newton-type methods is computationally expensive. Moreover, these methods can trap in undesirable local minima when the starting model lacks low-wavenumber part and the recorded data lack low-frequency content. In this paper, the Gauss-Newton (GN) method is modified to address these issues. We rewrite the GN system for multisoure multireceiver FWI in an equivalent matrix equation form whose solution is a diagonal matrix, instead of a vector in the standard system. Then we relax the diagonality constraint, lifting the search direction from a vector to a matrix. This relaxation is equivalent to introducing an extra degree of freedom in the subsurface offset axis for the search direction. Furthermore, it makes the Hessian matrix separable and easy to invert. The relaxed system is solved explicitly for computing the desired search direction, requiring only inversion of two small matrices that deblur the data residual matrix along the source and receiver dimensions. Application of the Extended GN (EGN) method to solve the extended-source FWI leads to an algorithm that has the advantages of both model extension and source extension. Numerical examples are presented showing robustness and stability of EGN algorithm for waveform inversion.

show abstract

Section: Computing δM From ∆Mmentioning

confidence: 99%

“…This idea comes from the fact that the reflection energy interfere constructively within the first Fresnel zone [16,22]. Yang et al [31] used a similar idea but they averaged the coefficients over a horizontal disk centered at the physical point.…”

Section: Computing δM From ∆Mmentioning

confidence: 99%

An extended Gauss-Newton method for full waveform inversion

Gholami¹

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…RTM is a seismic technique, which provides the visualization of the subsurface reflectivity using recorded seismic data. It is highly sensitivity to velocity errors, which result in a defocused and incoherent migration image [41]. Hence, RTM technique has been widely used as a quality-control (QC) tool in seismic imaging [42].…”

Section: ) Low-resolution Inversionmentioning

confidence: 99%

Multiscale Data-driven Seismic Full-waveform Inversion with Field Data Study

Feng,

Lin,

Wohlberg

2021

Preprint

View full text Add to dashboard Cite

Seismic full-waveform inversion (FWI), which applies iterative methods to estimate high-resolution subsurface detail from seismograms, is a powerful imaging technique in exploration geophysics. In recent years the computational cost of FWI has grown exponentially due to the increasing size and resolution of seismic data. Moreover, it is a non-convex problem, and can become stuck in a local minima due to the limited accuracy of the initial velocity maps, the absence of low frequencies in the measurements, the presence of noise, and the approximate modeling of the wave-physics complexity. To overcome these computational issues, we develop a multiscale datadriven FWI method based on the fully convolutional network (FCN). In preparing the training data, we first develop a realtime style transform method to create a large set of physically realistic subsurface velocity maps from natural images. We then develop two convolutional neural networks with encoder-decoder structure to reconstruct the low-and high-frequency components of the subsurface velocity maps, respectively. To validate the performance of our new data-driven inversion method and the effectiveness of the synthesized training set, we compare it with conventional physics-based waveform inversion approaches using both synthetic and field data. These numerical results demonstrate that, once our model is fully trained, it can significantly reduce the computation time, and yield more accurate subsurface velocity map in comparison with conventional FWI.

show abstract

“…Chen et al [34] inverted P-and S-wave impedance perturbations with pseudo-Hessian preconditioning in ELSRTM. In 2019, Yang et al [35] utilized an LSRTM scheme based on the subsurface offset domain extended imaging condition to obtain a better resolution when large velocity errors exist. In 2021, Zhong et al [36] developed a new scheme based on the elastic velocity−stress equation in isotropic media, and inverted reflectivities of Pand S-wave velocity.…”

Section: Introductionmentioning

confidence: 99%

Elastic Correlative Least-Squares Reverse Time Migration Based on Wave Mode Decomposition

Zheng

Liu

et al. 2022

Processes

View full text Add to dashboard Cite

The conventional elastic least-squares reverse time migration (LSRTM) generally inverts the parameter perturbation of the model rather than the reflectivity of reflected P- and S-modes, which leads to difficulty in directly interpreting the physical properties of the subsurface media. However, an accurate velocity model that is needed by the separation of seismic records of conventional LSRTM is usually unavailable in real data, which limits its application. In this study, we introduce a new practical correlative LSRTM (CLSRTM) scheme based on wave mode decomposition without amplitude and phase distortion, which frees from separation of seismic records. In this study, we deduced the migration and the de-migration operators using the decoupled P- and S-wave equations in heterogeneous media, which needs no extra wavefield decomposition in simulated data. To accelerate the convergence and improve the efficiency of the inversion, we adopted an analytical step-length formula that can be incidentally computed during the necessary de-migration process and the L-BFGS algorithm. Two numerical examples demonstrate that the proposed method can compensate the energy of deep structures, and generate clear images with balanced amplitudes and enhanced resolution even for the fault structures beneath the salt dome.

show abstract

Least-squares reverse time migration in the presence of velocity errors

Cited by 24 publications

References 57 publications

An extended Gauss-Newton method for full waveform inversion

An extended Gauss-Newton method for full waveform inversion

Multiscale Data-driven Seismic Full-waveform Inversion with Field Data Study

Elastic Correlative Least-Squares Reverse Time Migration Based on Wave Mode Decomposition

Contact Info

Product

Resources

About