Full Waveform Inversion Objective Research Articles

An extended Gauss-Newton method for full-waveform inversion

Full-waveform inversion (FWI) is a large-scale nonlinear ill-posed problem for which computationally expensive Newton-type methods can become trapped in undesirable local minima, particularly when the initial model lacks a low-wavenumber component and the recorded data lack low-frequency content. A modification to the Gauss-Newton (GN) method is developed to address these issues. The standard GN system for multisource multireceiver FWI is reformulated into an equivalent matrix equation form, with the solution becoming a diagonal matrix rather than a vector as in the standard system. The search direction is transformed from a vector to a matrix by relaxing the diagonality constraint, effectively adding a degree of freedom to the subsurface offset axis. The relaxed system can be explicitly solved with only the inversion of two small matrices that deblur the data residual matrix along the source and receiver dimensions, which simplifies the inversion of the Hessian matrix. When used to solve the extended-source FWI objective function, the extended GN (EGN) method integrates the benefits of model and source extension. The EGN method effectively combines the computational effectiveness of the reduced FWI method with the robustness characteristics of extended formulations and offers a promising solution for addressing the challenges of FWI. It bridges the gap between these extended formulations and the reduced FWI method, enhancing inversion robustness while maintaining computational efficiency. The robustness and stability of the EGN algorithm for waveform inversion are demonstrated numerically.

Robust joint adaptive multiparameter waveform inversion with attenuation compensation in viscoacoustic media

Full-waveform inversion (FWI) has been proven as an effective method to estimate subsurface parameters by iteratively reducing the data residual between the predictions and the observations. Nevertheless, FWI is greatly dependent on the initial model, and a poor initial model will lead to an incorrect solution. Furthermore, owing to the anelasticity of the earth, seismic waves will attenuate during propagation, which results in an attenuated gradient and makes the convergence rate of FWI even worse in viscoacoustic medium. Commonly, multiparameter (e.g. velocity and Q) waveform inversion can be used to mitigate these problems. Benefiting from the theory of Q-compensated wavefield propagation, we develop a Q-compensated joint multiparameter waveform inversion method to weaken the nonlinearity of the FWI objective function, which enables it to cope with challenges related with attenuation-induced gradient energy loss and cycle skipping simultaneously. We refer to our Q-compensated joint multiparameter FWI scheme as QJMFWI. The main contributions of QJMFWI are as follows: (1) given the difficulty associated with the estimation of velocity and Q simultaneously in viscoacoustic media, QJMFWI provides a straightforward waveform inversion method for velocity and Q model construction, by which we can obtain velocity and Q information with improved accuracy and resolution; and (2) compared with conventional FWI methods, QJMFWI relaxes the requirement for a good initial velocity and Q model, which can avoid trapping into local minima. Numerical and field data examples demonstrate that QJMFWI is an effective method to invert for accurate subsurface parameters in viscoacoustic media.

A Graph-Space Optimal Transport Approach Based on Kaniadakis κ-Gaussian Distribution for Inverse Problems Related to Wave Propagation.

Data-centric inverse problems are a process of inferring physical attributes from indirect measurements. Full-waveform inversion (FWI) is a non-linear inverse problem that attempts to obtain a quantitative physical model by comparing the wave equation solution with observed data, optimizing an objective function. However, the FWI is strenuously dependent on a robust objective function, especially for dealing with cycle-skipping issues and non-Gaussian noises in the dataset. In this work, we present an objective function based on the Kaniadakis κ-Gaussian distribution and the optimal transport (OT) theory to mitigate non-Gaussian noise effects and phase ambiguity concerns that cause cycle skipping. We construct the κ-objective function using the probabilistic maximum likelihood procedure and include it within a well-posed version of the original OT formulation, known as the Kantorovich-Rubinstein metric. We represent the data in the graph space to satisfy the probability axioms required by the Kantorovich-Rubinstein framework. We call our proposal the κ-Graph-Space Optimal Transport FWI (κ-GSOT-FWI). The results suggest that the κ-GSOT-FWI is an effective procedure to circumvent the effects of non-Gaussian noise and cycle-skipping problems. They also show that the Kaniadakis κ-statistics significantly improve the FWI objective function convergence, resulting in higher-resolution models than classical techniques, especially when κ=0.6.

Elastic FWI Without Low Frequencies Based on n-th Power Operation and Convolved Wavefields

The elastic full-waveform inversion (FWI) can use the recorded multi-component seismic data to construct high-precision multi-parameter models of the subsurface media such as P- and S-wave velocity models. However, due to the reasons such as data quality and algorithm limitations, there are still many problems in the promotion and application of elastic FWI method. Aiming at alleviating the influence of low-frequency data absence on the inversion results, we propose a robust elastic FWI method based on the n-th power operation. The n-th power of the seismic data can compress the time-domain waveform and expand its frequency-band. The FWI objective function constructed using the n-th power wavefields shows better convexity. By successively lowering the power during the inversion, we can realize a new multiscale FWI strategy, which is also a data-domain layer-stripping strategy. Seismic data will be more sensitive to the source wavelet errors after the n-th power operation. To mitigate this problem, we propose a robust objective function for elastic FWI using the n-th power operation and the convolved wavefields. Finally, the validity of the method is verified by numerical examples.

Applied 3D salt body reconstruction using shape optimization with level sets

As oil and gas extraction becomes more advanced, deep-water exploration becomes increasingly focused on imaging near or under complex salt geology, which necessitates detailed velocity models with strong contrast interfaces. These interfaces can be elegantly tracked using the level sets of an implicit surface. One can invert for the velocity model that best fits the recorded data in a full-waveform inversion (FWI) style objective function by reparameterizing the model in terms of an implicit surface representation of the salt interface. With this parameterization of the FWI objective function, we find the Hessian and solve a conjugate gradient system for the Newton step at every nonlinear iteration. We sparsify the representation of the implicit surface using radial basis functions, which can hasten convergence of the inner inversion by reducing the number of model parameters. We have developed a guided inversion approach that embeds information about the certainty of different salt boundary regions by the initialization of the implicit surface slope at the salt interface. This can help guide the inversion away from perceived local minima. The results of testing this inversion workflow on a 3D Gulf of Mexico data set show that it can be a useful tool for refining salt models because the initial and final seismic images show clearer and more consistent features below the updated salt area.

Volume source-based extended waveform inversion

Full-waveform inversion (FWI) faces the persistent challenge of cycle skipping, which can result in stagnation of the iterative methods at uninformative models with poor data fit. Extended reformulations of FWI avoid cycle skipping through adding auxiliary parameters to the model so that a good data fit can be maintained throughout the inversion process. The volume-based matched source waveform inversion algorithm introduces source parameters by relaxing the location constraint of source energy: It is permitted to spread in space, while being strictly localized at time [Formula: see text]. The extent of source energy spread is penalized by weighting the source energy with distance from the survey source location. For transmission data geometry (crosswell, diving wave, etc.) and transparent (nonreflecting) acoustic models, this penalty function is stable with respect to the data-frequency content, unlike the standard FWI objective. We conjecture that the penalty function is actually convex over much larger region in model space than is the FWI objective. Several synthetic examples support this conjecture and suggest that the theoretical limitation to pure transmission is not necessary: The inversion method can converge to a solution of the inverse problem in the absence of low-frequency data from an inaccurate initial velocity model even when reflections and refractions are present in the data along with transmitted energy.

Full-waveform inversion, Part 3: Optimization

This tutorial is the third part of a full-waveform inversion (FWI) tutorial series with a step-by-step walkthrough of setting up forward and adjoint wave equations and building a basic FWI inversion framework. For discretizing and solving wave equations, we use Devito ( http://www.opesci.org/devito-public ), a Python-based domain-specific language for automated generation of finite-difference code ( Lange et al., 2016 ). The first two parts of this tutorial ( Louboutin et al., 2017 , 2018 ) demonstrated how to solve the acoustic wave equation for modeling seismic shot records and how to compute the gradient of the FWI objective function using the adjoint-state method. With these two key ingredients, we will now build an inversion framework that can be used to minimize the FWI least-squares objective function.

Full-waveform inversion, Part 2: Adjoint modeling

This is the second part of a three-part tutorial series on full-waveform inversion (FWI) in which we provide a step-by-step walk through of setting up forward and adjoint wave equation solvers and an optimization framework for inversion. In Part 1 ( Louboutin et al., 2017 ), we showed how to use Devito ( http://www.opesci.org/devito-public ) to set up and solve acoustic wave equations with (impulsive) seismic sources and sample wavefields at the receiver locations to forward model shot records. Here in Part 2, we will discuss how to set up and solve adjoint wave equations with Devito and, from that, how we can calculate gradients and function values of the FWI objective function.

Modified Gauss-Newton full-waveform inversion explained — Why sparsity-promoting updates do matter

Full-waveform inversion (FWI) can be formulated as a nonlinear least-squares optimization problem. This nonconvex problem can be computationally expensive because it requires repeated solutions of the wave equation. Randomized subsampling techniques allow us to work with small subsets of (monochromatic) source experiments, reducing the computational cost. However, this subsampling may weaken subsurface illumination or introduce subsampling-related incoherent artifacts. These subsampling-related artifacts — in conjunction with the desire to obtain high-fidelity inversion results — motivate us to come up with a technique to regularize this inversion problem. Following earlier work, we have taken advantage of the fact that curvelets represent subsurface models and model perturbations parsimoniously. At first impulse, promoting sparsity on the model directly seemed the most natural way to proceed, but we have determined that in certain cases it can be advantageous to promote sparsity on the Gauss-Newton updates instead. Although constraining the one norm of the descent directions did not change the underlying FWI objective, the constrained model updates remained descent directions, removed subsampling-related artifacts, and improved the overall inversion result. We have empirically observed this phenomenon in situations where the different model updates occurred at roughly the same locations in the curvelet domain. We have further investigated and analyzed this behavior, in which nonlinear inversions benefit from sparsity-promoting constraints on the updates, by means of a set of carefully selected examples including the phase retrieval problem and time-harmonic FWI. In all cases, we have observed a faster decay of the residual and model error as a function of the number of iterations.

Full model wavenumber inversion Petroleum keynote paper

The promise of full waveform inversion (FWI), granted we could (completely) simulate the conditions in which our field data were acquired, including an accurate representation of the physics involved, is a model of the Earth capable of generating synthetic data that resembles (fits) our field data; No deghosting, no filtering, in fact, no processing is theoretically required as the goal in this case is a model of the Earth, not a seismic section. However, we still have a long way to go, as we usually resort to approximate physics and many assumptions, and yet, we still fail to converge to that promise. The high nonlinearity of the inversion problem with the large number of model points necessary to properly represent the resolution of interest (or need) in our model are prime reasons for the ill convergence. The long wavelength components of the velocity model usually constrain the general geometrical behavior of the wavefield (the kinematics), observed in our data, while the short wavelength components are responsible for the scattering (the reflections themselves as events in our data). Since FWI is based on comparing the observed and modeled data, free of wavefield geometrical utilization, it usually requires that the long (and at depth and with complex media, the middle) wavelength components of the model be accurate enough to provide modeled data that is within a half cycle of the observed data. These long wavelength components are usually estimated from tomography or migration velocity analysis (MVA) methods. They, however, usually contain too low of a wavelength to fit the cycle skip criteria for all reflections in the data, especially those reflections corresponding to deeper reflectors. We are, specifically, missing the model wavenumbers necessary to help us transition from the geometrical features of the wavefield to the scattering ones. The analysis of the sensitivity of our conventional (and even FWI-in-mind enhanced) data to the model points show that such lack of middle wavenumbers is a serious problem at depth, and specifically the depths we tend care about in our industry. In this presentation, we investigate two potential solutions to this problem, beyond requiring low frequency to be acquired. To combat this problem we shift our focus from the data domain to the model domain in which we devise an approach to explicitly control the wavenumbers that we introduce to the model at different stages of the inversion. Such controls are admitted naturally by scattering angle filters. An explicit control on the model wavenumbers provided by the scattering angle of the FWI gradients can help us maneuver model wavenumber gap. This is especially true in anisotropic media where such filters are applicable to the individual parameter models necessary to represent such the anisotropic model; A feature not accessible though data domain decimation and data hierarchical implementations, as all parameters share the same data. Though the physics involved in creating the data and the obvious acquisition limitations will eventually impose bounds on the model wavenumbers we may be able to extract, a proper integration of image domain analysis to the FWI objective will help us widen the model wavenumber spectrum that we can extract from the data. Thus, the combination of scattering angle filtering and an objective that utilizes MVA and FWI are at the heart of making FWI work, and hopefully help us converge to it's promise. During this presentation, I will share many examples that demonstrate the assertions made in this summary.

Alternative waveform inversion for surface wave analysis in 2-D media

In the context of near surface seismic imaging (a few hundreds of metres), we propose an alternative approach for inversion of surface waves in 2-D media with laterally varying velocities. It is based on Full Waveform Inversion (FWI) but using an alternative objective function formulated in the frequency–wavenumber f − k domain. The classical FWI objective function suffers from severe local minima problems in the presence of surface waves. It thus requires a very accurate initial model. The proposed objective function is similar to the one used in classical surface wave analysis. In this approach, the data are first split using sliding windows in the time–space t − x domain. For each window, the amplitude of the f − k spectrum is computed. The objective function measures the least-squares misfit between the amplitude of observed and modelled 2-D Fourier transformed data sets. We call this formulation the windowed-amplitude waveform inversion (w-AWI). The w-AWI objective function reduces some local minima problems as shown here through numerical examples. The global minimum basin is wider in the w-AWI approach than in FWI. Synthetic examples show that w-AWI may achieve convergence if the lowest data frequency content is twice higher than the one needed by FWI. For elastic inversion, w-AWI can be used to reconstruct a velocity model explaining surface waves. This surface wave inversion procedure can be used to retrieve near-surface model parameters in lateral-varying media