Full Waveform Inversion Research Articles

A modified symplectic discontinuous Galerkin method for acoustic and elastic wave simulations

Numerically solving seismic wave equations is vital to large-scale forward modeling and full waveform inversion. In this paper, a new modified symplectic discontinuous Galerkin (MSDG) method is proposed to solve the acoustic and elastic equations. The MSDG method employs a symmetric interior penalty Galerkin formulation as the space discretization. The time discretization is based on a modified symplectic partitioned Runge–Kutta scheme with minimized phase error. Thus, the MSDG method has the advantages of high accuracy, being flexible to deal with complex geometric boundaries and internal structures, and stable for long time simulations. The numerical stability conditions, dispersion and dissipation are investigated in detail for the MSDG method. To validate the method, we carry out several numerical examples for solving the acoustic and elastic wave equations in various media. The numerical results show that the MSDG method can effectively suppress the numerical dispersion and is suitable for wavefield simulations.

Salt reconstruction in full-waveform inversion using topology optimization techniques

SUMMARY With the increasing advances in the oil and gas industry, seismic imaging near or under salt structures has become an important point in deep-water exploration. Detailed velocity models of these areas are particularly interesting not only to characterize hydrocarbon reservoirs but also to identify potential sites for hydrogen and carbon dioxide storage in offshore salt caverns. Thus, we study the full-waveform inversion for the salt reconstruction in acoustic media with constant density considering the time-harmonic wave propagation in a finite element formulation using the topology optimization (TO) method. This problem is challenging due to the strong velocity contrast between salt bodies and the sedimentary background, in addition to the lack of low-frequency data and the inherent ill-posedness of the inverse problem. In this context, we incorporate techniques from the TO field, usually used in design applications, to overcome or reduce these known problems. We initially defined the squared slowness as a combination of two fields, one related to the salt shape and the other to the background. An interpolation rule based on the solid isotropic material with penalization method, combined with filtering and projection schemes, is used to find the shape of the salt bodies with increased sharpness interfaces. A Helmholtz-type filter is applied to modify the gradient aimed to regularize the problem and provide a more stable way for the salt shape to evolve during the inversion process. In particular, we demonstrate that the proposed approach may be relevant for reconstructing media with salt bodies when a suitable starting model is unavailable, and sharp interfaces are required. In addition, we present inversion results from synthetic data generated by a variable density model to demonstrate the approach capability when subjected to a reconstruction application.

Vertical Seismic-Profile Data Local Full-Waveform Inversion Based on Marchenko Redatuming

Local full-waveform inversion (FWI) methods use redatumed seismic responses of virtual receivers within the subsurface to build the local objective function based on the convolution-type representation theorem. The Marchenko method is widely used to obtain the redatumed data. The method only requires a smoothed velocity model with correct kinematic characteristics of seismic responses for redatuming of the single-sided reflection data. However, the standard Marchenko method is insensitive to lateral propagation of the wavefield. By injecting the standard Marchenko redatumed wavefield along the boundary of the target, the local wavefield propagation modeling produces errors, which affects the accuracy of the local FWI. In this paper, a method to obtain more accurate Green’s functions is proposed by incorporating vertical seismic profile data (VSP) into the calculation process of the Marchenko source-receiver redatuming. This method allows one to obtain the accurate laterally propagating waveform, resulting in a significant improvement of lateral resolution. The proposed method is applied to a benchmark model dataset and compared with the local FWI based on standard Marchenko redatuming.

Elastic and anelastic adjoint tomography with and full Hessian kernels

SUMMARYThe elastic and anelastic structures of the Earth offer fundamental constraints for understanding its physical and chemical properties. Deciphering small variations in the velocity and amplitude of seismic waves can be challenging. Advanced approaches such as full-waveform inversion (FWI) can be useful. We rewrite the anelastic Fréchet kernel expression of Fichtner & van Driel using the displacement–stress formulation. We then derive the full Hessian kernel expression for viscoelastic properties. In these formulations, the anelastic Fréchet kernels are computed by the forward strain and a shift of the adjoint strain. This is complementary to the quality factor Q (i.e., inverse attenuation) Fréchet kernel expressions of Fichtner & van Driel that are explicit for the velocity–stress formulation. To reduce disk space and I/O requirements for computing the full Hessian kernels, the elastic full Hessian kernels are computed on the fly, while the full Hessian kernels for Q are computed by a combination of the on-the-fly approach with the parsimonious storage method. Applications of the Fréchet and full Hessian kernels for adjoint tomography are presented for two synthetic 2-D models, including an idealized model with rectangular anomalies and a model that approximates a subduction zone, and one synthetic 3-D model with an idealized geometry. The calculation of the full Hessian kernel approximately doubles the computationally cost per iteration of the inversion; however, the reduced number of iterations and fewer frequency stages required to achieve the same level of convergence make it overall computationally less expensive than the classical Limited-memory Broyden–Fletcher–Goldfarb–Shanno (L-BFGS) FWI for the 2-D elastic tested models. We find that the use of full Hessian kernels provides comparable results to the L-BFGS inversion using the improved anelastic Fréchet kernels for the 2-D anelastic models tested for the frequency stage up to 0.5 Hz. Given the computational expense of the Q full Hessian kernel calculation, it is not advantageous to use it in Q inversions at this time until further improvements are made. For the 3-D elastic inversion of the tested model, the full Hessian kernel provides similar image quality to the L-BFGS inversion for the frequency stage up to 0.1 Hz. We observe an improved convergence rate for the full Hessian kernel inversion in comparison to L-BFGS at a higher frequency stage, 0.1–0.2 Hz, and we speculate that at higher frequency stages the use of full Hessian kernels may be more computationally advantageous than the classical L-BFGS for the tested models. Finally, we perform 3-D elastic and Q L-BFGS inversions simultaneously using the rederived Q kernels, which can reduce the computational cost of the inversion by about 1/3 when compared to the classical anelastic adjoint tomography using the additionally defined adjoint source. The recovered Q model is smeared when compared to the recovered elastic model at the investigation frequencies up to 0.5 Hz. Q inversion remains challenging and requires further work. The 2-D and 3-D full Hessian kernels may be used for other purposes for instance resolution analysis in addition to the inversions.

Enhancing salt model resolution and subsalt imaging with elastic FWI

Full-waveform inversion (FWI) has become the centerpiece of velocity model building (VMB) in seismic data processing in recent years. It has proven capable of significantly improving the velocity model and, thus, the migration image for different acquisition types and geologic settings, including complex environments such as salt. With the advent of FWI imaging, the scope of FWI applications has extended further from VMB into the imaging landscape. However, current FWI applications in the industry prevalently employ the acoustic approximation. One common problem of acoustic FWI (A-FWI) is the apparent salt halos at the salt-sediment interface in the resulting FWI velocity and FWI image, the presence of which hinders direct interpretation and imaging focusing around salt bodies. With synthetic and field data examples, we demonstrate that this salt halo is caused mainly by the large mismatch between the elastic recorded data and the acoustic modeled data, particularly at middle to long offsets. To overcome limitations imposed by acoustic assumptions, we developed an elastic FWI (E-FWI) algorithm that combines an elastic modeling engine with the time-lag cost function, which we call elastic time-lag FWI (E-TLFWI). With a more accurate modeling engine, E-TLFWI significantly reduces the salt halo observed in its acoustic counterpart. However, we also observe that the images migrated using the A-FWI and E-FWI velocity models remain similar overall, with some slight improvements around and beneath salt boundaries, particularly near steep salt flanks, as a result of the reduced salt halo. By contrast, FWI images derived from E-TLFWI show considerable benefits over those from acoustic time-lag FWI, such as improved event focusing, better structural continuity, and higher signal-to-noise ratio. The sharpened salt boundaries and enhanced quality of the FWI images reveal the significant value of E-FWI and provide the justification for its greatly increased cost.

Waveform Inversion of Shallow Seismic data with Randomly Selected Sources

Applying waveform inversion with randomly selected sources (RS) increases the convergence rate of the optimization process within full-waveform inversion (FWI) workflows and reduces overall computational time. FWI has been shown to be a valuable addition to the existing geophysical methods for near-surface characterization. Accurate 3D modeling of the (visco) elastic wavefield allows to diminish assumptions about wave propagation and include surface- and body-wave based arrivals within the inversion workflow. This approach could result in reliable high-resolution subsurface models, but generally it comes at a high computational cost for each individual source modeling. Commonly, multiple sources are involved in the inversion process, which proportionally increases the resulting computations for a time-domain FWI, but seismic data with dense source/receiver coverage usually contain redundant information. This is especially true for seismic near-surface applications, where the number of recorded sources per wavelength of interest are normally excessive. The inversion performance was increased by randomly selecting a subset of sources at each FWI iteration. The method's effectiveness is obvious on a FWI near-surface void detection application. Synthetic 2D experiments for fixed and rolling spreads showed comparable results with fewer calculations. The best performance was achieved when a single random source was used for each inversion iteration. The effectiveness of the method was also evident on a shallow 3D field dataset collected in the Sonora Desert in western Arizona, where data were acquired over a 10-m deep void with known location, orientation, and dimensions.

Robust and efficient waveform-based velocity-model building by optimal transport in the pseudotime domain: Methodology

Full-waveform inversion (FWI) aims at a broadband reconstruction of the subsurface physical properties by fitting the entire recorded wavefield. In realistic exploration seismic surveys, however, conventional FWI often fails to retrieve the deep velocity model due to the limited penetration depth of diving waves. Joint FWI (JFWI) unifies reflection-waveform inversion and early arrival waveform inversion to reconstruct simultaneously the shallow and deep subsurface kinematics. However, several factors limit the appeal of JFWI velocity-model building: (1) conflict between fixed reflectivity and evolving kinematics, creating phase ambiguity at short offsets; (2) susceptibility to cycle skipping at mid-to-long offsets, thus reliance on the quality of the starting model; and (3) cost of building and updating the reflective model. We have developed a fully operational JFWI-based methodology that systematically addresses the aforementioned issues. JFWI is reformulated in the pseudotime domain, to enforce consistency between velocity and reflectivity in a cost-effective fashion, without repeated least-squares migrations. A JFWI graph-space optimal transport (GSOT) objective function is designed to avert cycle skipping, whereas nonuniqueness is mitigated at no extra cost by smoothing the velocity gradient along the structures extracted from the reflective model. A dedicated asymptotic-based preconditioner is developed for impedance waveform inversion, making it possible to obtain sharp and balanced reflective images in a fraction of the time. We determine that pseudotime GSOT-JFWI retrieves complex velocity macromodels from limited-offset data sets with minimal preprocessing, starting from noninformative initial solutions. Compared with depth-domain JFWI, the computing cost is reduced significantly, along with a simpler and less subjective design of data weighting and inversion strategy. Pseudotime GSOT-JFWI provides FWI with the necessary low wavenumbers to converge to the broadband model, reducing the need for accurate starting models, on the road to a fully waveform-based imaging workflow.

Introduction to this special section: Full-waveform inversion

The estimation of the parameters of a mathematical model by data-fitting procedures goes back more than 200 years. Mathematics historians seem to agree to credit C. F. Gauss for the introduction of this idea ( Gauss, 2011 ). In the field of exploration geophysics, Lailly (1983) and Tarantola (1984) were the first to propose the use of data-fitting techniques to estimate model parameters that control the propagation of waves in the subsurface. The concept of full-waveform inversion (FWI) was born.

Application of full-waveform inversion to land data: Case studies in onshore Mexico

Velocity model building and imaging for land surveys are often challenging due to near-surface complexity contaminating the reflection signal. Incorporating full-waveform inversion (FWI) in the velocity model building workflow for land surveys offers benefits not achieved with traditional model building tools, but it also brings difficulties. We have developed an effective model building workflow for land seismic data that incorporates dynamic matching FWI (DMFWI). DMFWI employs an objective function that uses multidimensional local windowed crosscorrelations between the dynamically matched version of observed and synthetic data. Dynamic matching de-emphasizes the impact of amplitudes, allowing the algorithm to focus on using kinematic information for velocity updates. The proposed workflow produces a geologically plausible and consistent model for data acquired with limited offsets. Refraction and reflection tomography may also be included in the workflow. The workflow is applied to onshore surveys in Mexico. Despite challenges of the near-surface geology and limitations of the acquisition parameters in the study areas, the proposed model building workflow successfully derives a high-resolution velocity model that significantly improves the migrated depth image.

2D frequency-domain finite-difference acoustic wave modeling using optimized perfectly matched layers

Seismic forward modeling is fundamental in sensitivity analysis and full-waveform inversions. In finite-difference acoustic wavefield simulation, the absorption boundary, especially the perfectly matched layer (PML), is widely used, but the setting of PML parameters is empirical. An optimized complex frequency-shift PML (CFS-PML) for the modeling of acoustic fields has been developed. It refines the selection of parameters for improving the artificial attenuation. The improved CFS-PML boundary condition is applied to a 2D frequency-domain acoustic wave simulation using an optimal 17-point finite-difference scheme. Numerical results are compared with a conventional nine-point scheme in terms of computational time and physical memory consumption. These tests indicate that our CFS-PML absorption boundary can effectively improve numerical accuracy without increasing the computational burden remarkably.

Computationally efficient deep learning-based image reconstruction for sound speed estimation in ultrasound computed tomography

Ultrasound computed tomography (USCT) has great potential for breast cancer imaging due to its ability to quantify acoustic tissue properties such as the speed of sound (SOS). Full-waveform inversion (FWI)-based image reconstruction methods can produce high spatial resolution SOS images, but these methods are computationally demanding, especially when formulated in the time-domain. Deep learning (DL)-based methods can be computationally efficient and provide high-quality object estimates as studied in various medical image reconstruction problems. However, there remains a further need to develop and assess the DL-based image reconstruction methods for USCT for use in medical imaging applications. Here, we present a supervised DL approach for USCT image reconstruction that is formulated as an image-to-image translation problem. Specifically, a low-resolution SOS image obtained via traveltime tomography and a high-resolution reflection tomography image are used as inputs to a U-Net, which is trained to estimate a high-resolution SOS map. These input images are complementary, providing both quantitative information about SOS values and information about tissue boundaries. Numerical studies that employ realistic numerical breast phantoms demonstrated that the proposed method produced high-quality SOS estimates based on image quality metrics such as NMSE, SSIM, and PSNR.

Full-waveform inversion in acoustic-elastic coupled media with irregular seafloor based on the generalized finite-difference method

The marine environment can be considered as a fluid-solid coupled media governed by the acoustic-elastic coupled equation (AECE). For complex fluid-solid coupled models with irregular seabeds, the well-known finite-difference method (FDM) often causes ripple-shaped scattering due to the rectangular grid. To overcome this problem, a meshless method (MM), called the generalized finite-difference method (GFDM), for irregular node distributions is developed and applied to AECE forward modeling by using a body-fitted node-generation method, which provides robust node distributions for GFDM. Compared with the FDM, our GFDM improves modeling accuracy and eliminates divergence. Furthermore, our MM was applied to an AECE-based elastic full-waveform inversion (EFWI). Owing to the well-matched irregular seabed, our EFWI can invert the accurate elastic model for ocean-bottom cable and pressure data. Numerical examples demonstrate that this method can efficiently improve the numerical accuracy and build an accurate velocity model even using noisy data, which is helpful for high-precision seismic exploration in the complex marine environment.

Free-surface elastic wave injection and wavefield reconstruction with applications to elastic reverse time migration

Accurately injecting and extrapolating multicomponent seismic data is a major step in elastic reverse time migration (RTM) and full-waveform inversion. The theoretical basis of this process is the seismic representation theorem. We propose a numerical approach based on the collocated-grid MacCormack finite-difference (FD) scheme to implement the injection and extrapolation. The MacCormack FD scheme defines wavefield parameters on the same grid. When dealing with the boundary condition in the representation theorem, it is more advantageous than the commonly used staggered-grid FD scheme which defines the stress and velocity on offset grids; thus, interpolations are required to link the boundary values to nearby nodes. In contrast, with the MacCormack scheme, the traction and particle velocity data can be directly injected into the boundary nodes as equivalent concentrated forces, force dipoles, and double couples. The accuracy of this approach is validated by comparing the regenerated wavefield with the original wavefield. Our primary purpose is to develop an injection method for elastic RTM, but this approach also can be used to numerically investigate other interesting issues in the elastic wave injection, including errors and artifacts caused by injection from unclosed boundaries, incomplete boundary values, and uncertain velocity models. In particular, seismic data are usually acquired on the ground surface where the traction-free condition is naturally maintained. We simulate a land acquisition scenario by generating a synthetic multicomponent data set on a traction-free surface. The reconstructed wavefields from velocity data are compared with the original forward-propagated waves to evaluate the accuracy of the boundary treatments. Two multicomponent elastic RTM examples are presented to demonstrate the applications of the proposed injection method. By combining the regenerated receiver-side wavefield and the vectorized imaging condition, the PP, PS, SP, and SS images are calculated.

Three-dimensional time-domain full-waveform inversion for ring-array-based ultrasound computed tomography

Ultrasound computed tomography (USCT) is a promising imaging technique for breast cancer diagnosis. Full-waveform inversion (FWI)-based image reconstruction methods based on the acoustic wave equation can produce quantitatively accurate images of acoustic properties with higher spatial resolution compared to ray-based methods. One common USCT design employs a circular ring-array of elevation-focused ultrasonic transducers that can be translated orthogonally to the imaging plane to achieve volumetric scanning. Currently, slice-by-slice (SBS) reconstruction methods are often used in this setting. Such methods use a 2D wave physic model to reconstruct a cross-sectional image for each position of the ring-array and vertically stack them together to render the 3D volume. However, this approach neglects the 3D wave propagation physics and transducer’s focusing properties, leading to out-of-plane scattering artifacts and degraded spatial resolution. To address this, a 3D time-domain FWI method that utilizes measurement data acquired at multiple adjacent ring-array locations is proposed. The reconstruction method was investigated using virtual imaging studies of ring-array-based USCT that employed realistic 3D numerical breast phantoms. The results demonstrated improved spatial resolution (both in-plane and off-plane) and reduced image artifacts compared to the SBS approach. Additionally, the impact of the number of ring-array locations on image quality is assessed.

Deep-Learning-Based Low-Frequency Reconstruction in Full-Waveform Inversion

Low frequencies are vital for full-waveform inversion (FWI) to retrieve long-scale features and reliable subsurface properties from seismic data. Unfortunately, low frequencies are missing because of limitations in seismic acquisition steps. Furthermore, there is no explicit expression for transforming high frequencies into low frequencies. Therefore, low-frequency reconstruction (LFR) is imperative. Recently developed deep-learning (DL)-based LFR methods are based on either 1D or 2D convolutional neural networks (CNNs), which cannot take full advantage of the information contained in 3D prestack seismic data. Therefore, we present a DL-based LFR approach in which high frequencies are transformed into low frequencies by training an approximately symmetric encoding-decoding-type bridge-shaped 3D CNN. Our motivation is that the 3D CNN can naturally exploit more information that can be effectively used to improve the LFR result. We designed a Hanning-based window for suppressing the Gibbs effect associated with the hard splitting of the low- and high-frequency data. We report the significance of the convolutional kernel size on the training stage convergence rate and the performance of CNN’s generalization ability. CNN with reasonably large kernel sizes has a large receptive field and is beneficial to long-wavelength LFR. Experiments indicate that our approach can accurately reconstruct low frequencies from bandlimited high frequencies. The results of 3D CNN are distinctly superior to those of 2D CNN in terms of precision and highly relevant low-frequency energy. FWI on synthetic data indicates that the DL-predicted low frequencies nearly resemble those of actual low frequencies, and the DL-predicted low frequencies are accurate enough to mitigate the FWI’s cycle-skipping problems. Codes and data of this work are shared via a public repository.

Shear wave elastography: From dispersion matching to full waveform inversion

Shear Wave Elastography (SWE) involves estimating mechanical properties through inversion, i.e., matching measured and simulated propagation characteristics of shear waves in the tissue. The accuracy of the estimated properties depends significantly on the specific characteristics/responses that are being matched. These could range from simple group velocity to dispersion curves and to full-wave response (particle velocity measurements). Using specific applications of arterial, liver, and tumor elstography, we illustrate that effective SWE is performed by resorting to an inversion approach, or combination of inversion approaches, guided by the underlying physics. To this end, we present inversion approaches ranging from matching dispersion characteristics to matching full waveform responses and provide rationale for choosing the appropriate technique(s) depending on the problem at hand.

Tutorial: Least Squares Migration and Full Waveform Imaging

A major underlying assumption of migration is that the input data are adequately sampled in terms of surface coverage. In addition, we hope that the subsurface is adequately illuminated, and that the migration algorithm itself is based on an acceptable numerical approximation of the wave equation. However, in general these assumptions and aspirations are never fully met, leading to amplitude imbalance and blurring of the output image. To some extent, this blurring and amplitude imbalance can be removed from the migrated data via application of some form of localised deconvolution, generally referred to as least-squares migration. This image modification can be performed in either the data or the image domain and can be achieved via an iterative or a single pass process, under the assumption that the velocity model is acceptable and that no coherent noise such as multiples, contaminates the input data. In this tutorial, we outline some of the various possible approaches to least-squares migration, and comment on the emerging ideas of adapting full waveform inversion to supplant some of the more expensive forms of least-squares image modification.

Is 3D frequency-domain FWI of full-azimuth/long-offset OBN data feasible? The Gorgon data FWI case study

Frequency-domain full-waveform inversion (FWI) is potentially amenable to efficient processing of full-azimuth long-offset stationary-recording seabed acquisition carried out with a sparse layout of ocean-bottom nodes (OBNs) and broadband sources because the inversion can be performed with a few discrete frequencies. However, computing the solution of the forward (boundary-value) problem efficiently in the frequency domain with linear algebra solvers remains a challenge for large computational domains involving tens to hundreds of millions of parameters. We illustrate the feasibility of 3D frequency-domain FWI with a subset of the 2015/2016 Gorgon OBN data set in the North West Shelf, Australia. We solve the forward problem with the massively parallel multifrontal direct solver MUMPS, which includes four key features to reach high computational efficiency: an efficient parallelism combining message-passing interface and multithreading, block low-rank compression, mixed-precision arithmetic, and efficient processing of sparse sources. The Gorgon subdata set involves 650 OBNs that are processed as reciprocal sources and 400,000 sources. Monoparameter FWI for vertical wavespeed is performed in the viscoacoustic vertically transverse isotropic approximation with a classical frequency continuation approach proceeding from a starting frequency of 1.7 Hz to a final frequency of 13 Hz. The target covers an area ranging from 260 km2 (frequency ≥ 8.5 Hz) to 705 km2 (frequency ≤ 8.5 Hz) for a maximum depth of 8 km. Compared to the starting model, FWI dramatically improves the reconstruction of the bounding faults of the Gorgon horst at reservoir depths as well as several intrahorst faults and several horizons of the Mungaroo Formation down to a depth of 7 km. Seismic modeling reveals a good kinematic agreement between recorded and simulated data, but amplitude mismatches between the recorded and simulated reflection from the reservoir suggest elastic effects. Therefore, future works involve multiparameter reconstruction for density and attenuation before considering elastic FWI from hydrophone and geophone data.

Application of Q-FWI-tomography and least-squares migration to improve seismic resolution in Tengiz oil field

Seismic image deterioration is a major problem for structures under complex overburdens. In areas under major faults and salt bodies, it is common to observe amplitude washout zones, poor structural definition, and relatively strong coherent and incoherent seismic noise. We developed a data-driven workflow of Q least-squares migration to mitigate these problems by enhancing the horizontal and vertical resolutions of seismic images. The workflow includes accurate velocity model building with joint full-waveform inversion and tomography to maximize the stacking power of P-wave primaries, Q tomography that balances weak-amplitude washout zones and minimizes swing noises, and least-squares migration with the aid of a constraint map that yields higher-resolution structural images. Its application on a wide-azimuth data set from the Tengiz oil field demonstrates the effective mitigation of fault shadow issues with an overall improvement of image quality. In addition, the uplift in focusing of diffracted energies shows promising improvements in enhancing seismic mega-amplitude events. Thus, the proposed method greatly increases the fidelity of seismic attribute analysis when compared with conventional vintage processing.

Acquisition and near-surface impacts on VSP mini-batch FWI and RTM imaging in desert environment

The SEG Advanced Modeling (SEAM) Arid benchmark model was designed to simulate an extremely heterogeneous low-velocity near surface (NS), which is typical of desert environments and typically not well characterized or imaged. Imaging of land seismic data is highly sensitive to errors in the NS velocity model. Vertical seismic profiling (VSP) partly alleviates the impact of the NS as the receivers are located at depth in the borehole. Deep learning (DL) offers a flexible optimization framework for full-waveform inversion (FWI), often outperforming typically used optimization methods. We investigate the quality of images that can be obtained from SEAM Arid VSP data by acoustic mini-batch reverse time migration (RTM) and full-waveform imaging. First, we focus on the effects of seismic vibrator and receiver array positioning and imperfect knowledge of the NS model when inverting 2D acoustic data. FWI imaging expectedly and consistently outperforms RTM in our tests. We find that the acquisition density is critical for RTM imaging and less so for FWI, while NS model accuracy is critical for FWI and has less effect on RTM imaging. Distributed acoustic sensing along the full length of the well provides noticeable improvement over a limited aperture array of geophones in imaging deep targets in both RTM and FWI imaging scenarios. Finally, we compare DL-based FWI imaging with inverse scattering RTM using the upgoing wavefield from the original SEAM data. Use of significantly more realistic 3D elastic physics for the simulated data generation and simple 2D acoustic inversion engine makes our inverse problem more realistic. We observe that FWI imaging in this case produces an image with fewer artifacts.