Searcharxiv⌕ Search

arXiv subjects

Tariq Alkhalifah

Publications and source records attributed to Tariq Alkhalifah.

At least 19 recordsLinked to original sources
arXiv2026

Shallow-to-deep velocity model building via diffusion models-Part I: Method and Proof of concept

Seismic velocity model building (VMB) is fundamental for understanding subsurface structures. Traditional methods demand high-quality starting models and, also, remain limited in resolution in coverage and computationally intensive. Recent generative diffusion model-based approaches capture statistical priors to support traditional inversion methods, but these approaches do not account for the top to bottom progression of information (layer stripping) involved in surface recorded data, where deep velocity information depends on the shallow. To address this issue, we propose a depth-progressive diffusion framework that constructs velocity models incrementally from shallow to deep by propagating prior information. Our method trains on paired shallow-deep velocity patches with variable overlap and explicit depth encoding, integrating multiple geophysical constraints including well logs and seismic images (representing structural information). During inference, we synthesize overlapping depth slices using a progressive algorithm and merge them with Gaussian-weighted blending to eliminate boundary artifacts. This approach leverages both learned geological distributions and observed shallow priors while providing uncertainty quantification. Extensive numerical experiments on in-distribution tests and an out-of-distribution test demonstrate excellent VMB accuracy with a strong correlation between predicted uncertainty and actual errors. As a proof of concept, this part I employs idealized structural constraints derived from vertical reflectivity to validate the methodological framework. The companion paper (Part II) extends the approach to realistic structural constraints relying on migrated images with field data applications.

physics.geo-ph↗
arXiv2026

Shallow-to-deep velocity model building via diffusion models-Part II: Realistic scenarios

Full-waveform inversion (FWI) requires accurate initial velocity models to avoid cycle-skipping, but constructing such models remains challenging in practice. Building on the depth-progressive diffusion framework introduced in Part~I, which relied on idealized reflectivity constraints, this work adapts the methodology to realistic exploration scenarios. We replace perfect structural information with migration-derived attributes extracted from seismic images, and introduce smooth background velocity models from tomography as additional conditioning inputs. The framework jointly leverages background/migration velocity, migrated structural information, and sparse well measurements to synthesize high-resolution velocity models through depth-progressive generation. Validation on synthetic examples demonstrates superior accuracy compared to conventional interpolation and alternative deep learning methods, with generated models successfully initializing FWI and mitigating cycle-skipping even in complex geological structures. Field data confirms practical applicability: despite training on synthetic data, the method generalizes effectively to field conditions, producing velocity models with synthetic data response that nearly match observed seismic data. As a result, this framework establishes a practical pathway to deploy generative diffusion models for velocity model building under realistic constraints.

physics.geo-ph↗
arXiv2026

Post-FWI Injection of Learned Priors Using a Flow Matching Model

Full Waveform Inversion (FWI) is a powerful tool for subsurface velocity reconstruction but remains highly ill-posed, sensitive to acquisition limitations, often requiring some form of regularization to reduce artifacts and enhance resolution. While recent developments have shown that generative models can inject learned priors directly into the FWI optimization process, such approaches typically require additional, computationally expensive inversion iterations. In this study, we propose a post-FWI refinement strategy based on a Flow Matching (FM) generative model, which leverages learned geological priors without re-running FWI. The method guides the deterministic generative process using the FWI result, as well as well logs, if available. Synthetic and field data experiments demonstrate that we can inject well information and our geological expectations (prior) into the provided FWI result, and thus, we can effectively enhance its resolution and geological quality. In fact, the well prior even managed to alter the model depth to fit the well information, which is a form of correcting for depth misties.

physics.geo-ph↗
arXiv2026

Estimation of Elastic Parameters with Guidance-based Diffusion model

Elastic parameters are fundamental rock properties for reservoir characterization, but their reliable estimation from angle-stack seismic data remains challenging due to strong nonlinearity and imperfect physical modeling. Conventional deterministic approaches based on linearized Zoeppritz approximations yield a single point estimate and cannot quantify solution uncertainty, while probabilistic methods are computationally expensive. To address these limitations, we present a workflow for elastic parameter inversion from angle-stack seismic data using a guided diffusion model as an implicit prior over the joint distribution of P-wave velocity, S-wave velocity, and density. The diffusion model is trained in an unsupervised manner on benchmark datasets and well-log-derived synthetic models, learning the non-Gaussian statistical coupling among the three elastic parameters. For guidance, we employ Diffusion Posterior Sampling (DPS), which approximates the likelihood function through a forward operator based on the Aki-Richards approximation and injects data-consistency gradient corrections at each reverse diffusion step, sampling from a posterior conditioned on the misfit between observed and modeled angle-stack data. We evaluate the framework on two datasets: the 2D Otway synthetic elastic model and field data from the Poseidon field, NW Shelf, Browse Basin, Australia, comparing it against two baselines: LSQR least-squares inversion and ADMM-based inversion with total variation regularization. Quantitative comparisons confirm that the diffusion-based framework recovers sharper lithological contrasts and geologically more realistic elastic profiles. Uncertainty quantification is achieved by generating multiple independent posterior realizations through repeated reverse diffusion runs, producing spatially resolved uncertainty maps for each elastic parameter.

physics.geo-ph↗
arXiv2026

Velocity Model Building and Editing with Guided Denoising Diffusion Implicit Models

Velocity-model building is a fundamental component of seismic imaging, yet it remains a challenging inverse problem due to limited data coverage, nonlinearity, and the need to integrate heterogeneous information such as well logs. We introduce a unified framework for velocity-model editing and full velocity-model building that combines learned diffusion priors with structurally preconditioned inverse formulations. A diffusion model trained on high-resolution synthetic velocity examples provides a data-driven prior that is exploited through Denoising Diffusion Implicit Model (DDIM) inversion and guided sampling. For localized editing, the diffusion prior is coupled with a structurally preconditioned Tikhonov well-matching inversion, enabling controlled modification of selected regions while preserving global consistency. For full velocity-model building, we formulate a well-matching inverse problem augmented with imaging-based regularization and solve it using conventional least-squares, the proposed DDIM-guided method, and Diffusion Posterior Sampling (DPS). Synthetic experiments demonstrate that diffusion-based approaches recover sharper and more realistic velocity structures than classical inversion. Field-data applications on the Viking Graben dataset confirm robustness under realistic acquisition conditions. An ablation study highlights the critical role of structural slope guidance in inversion performance. Overall, the proposed framework bridges inverse problems and generative modeling, offering a flexible approach for practical seismic imaging workflows.

physics.geo-ph↗
arXiv2026

Joint Velocity Slope Diffusion Prior for Structurally Constrained Velocity Model Building

High-resolution velocity models are crucial for reservoir characterization and subsurface delineation. However, the band limited nature of our surface recorded data limits resolution. Utilizing well measurements to enhance the resolution of our subsurface models is an important objective. To this end, we present a diffusion-guided framework for structurally preconditioned velocity-model reconstruction from sparse well-log information. The proposed approach combines plane-wave PDE regularization, structurally preconditioned inversion, and measurement-guided diffusion posterior sampling within a unified formulation. Local structural slopes estimated through plane-wave destruction are used both to propagate well information along geological dip directions and to guide the diffusion sampling process through a joint velocity--slope generative prior. Numerical experiments on the Volve synthetic model and the Viking Graben field dataset demonstrate that the proposed framework improves structural continuity, lateral consistency, and geological realism compared with conventional structurally preconditioned inversion approaches while maintaining computationally practical inference through DDIM sampling.

physics.geo-ph↗
arXiv2026

Generative wave propagator

Seismic wavefield simulation is fundamental to seismology, but conventional finite-difference (FD) methods remain limited by numerical dispersion and stability constraints, which often require dense spatial grids and small time steps and thereby severely limit the effectiveness of iterative inversion workflows. We introduce a conditional diffusion-based wavefield propagator that advances seismic wavefields recursively from one time step to the next. Instead of learning an unconditional data distribution of wavefield evolution, the model is conditioned by a short history of recent wavefield time steps (snapshots), the velocity model, and the wavefield time step index, allowing it to represent the conditional transition between adjacent physical states. By training the network to directly predict the clean next wavefield snapshot, this strong physical conditioning makes it possible to replace the iterative reverse diffusion process with a single network evaluation for each predicted snapshot. To improve stability over long recursive rollouts, we further introduce a causal time-weighted loss, in which adaptive weights, accumulated as exponential moving averages of per-snapshot training errors, emphasize training directions that are consistent with the forward propagation sequence and reduce the amplification of one-step prediction errors. Because the learned propagator is tied to the temporal spacing of the training snapshots rather than to the FD stability limit, it can advance the wavefield using a physical time step ten times larger than that required by the underlying solver. Experiments on the Overthrust, SEG/EAGE, and Marmousi models show that the proposed method accurately reproduces wavefield snapshots and shot gathers and achieves an end-to-end speedup of 2.17 x over a GPU-accelerated tenth-order staggered-grid FD implementation under matched hardware conditions.

physics.geo-ph↗
arXiv2026

Mitigating cycle skipping in full waveform inversion using max-pooling-based approximate envelope and shot patching

Full waveform inversion (FWI) can produce accurate subsurface velocity models. However, the lack of sufficiently low-frequency content in field data often causes cycle skipping and traps the inversion in local minima. The Hilbert-transform envelope (HTE) provides a low-frequency representation that helps mitigate cycle skipping, but it may be insufficient when the initial velocity model is highly inaccurate. To further enhance low-frequency information and reduce dependence on the initial model, we compute an approximate envelope using a sequence of 2D max-pooling operations. Compared with HTE, the resulting max-pooling-based approximate envelope (MPBAE) contains richer low-frequency components and better mitigates cycle skipping. We further combine the MPBAE loss with a shot patching strategy and exploit the inherent normalization property of the Euclidean loss to formulate the MPBAEP loss, in which each shot gather is divided into localized patches for misfit evaluation. This introduces local adjoint-source energy balancing, as the adjoint source associated with the Euclidean loss exhibits a normalization effect within each local region, thereby improving gradient balance and accelerating convergence. Numerical experiments on synthetic and field data demonstrate that MPBAE-FWI significantly outperforms HTE-FWI when the initial model is poor, while MPBAEP-FWI further improves inversion accuracy.

physics.geo-ph↗
arXiv2026

A Green-Integral-Constrained Neural Solver with Stochastic Physics-Informed Regularization

Standard physics-informed neural networks (PINNs) struggle to simulate highly oscillatory Helmholtz solutions in heterogeneous media because pointwise minimization of second-order PDE residuals is computationally expensive, biased toward smooth solutions, and requires artificial absorbing boundary layers to restrict the solution. To overcome these challenges, we introduce a Green-Integral (GI) neural solver for the acoustic Helmholtz equation. It departs from the PDE-residual-based formulation by enforcing wave physics through an integral representation that imposes a nonlocal constraint. Oscillatory behavior and outgoing radiation are encoded directly through the integral kernel, eliminating second-order spatial derivatives and enforcing physical solutions without additional boundary layers. Theoretically, optimizing this GI loss via a neural network acts as a spectrally tuned preconditioned iteration, enabling convergence in heterogeneous media where the classical Born series diverges. By exploiting FFT-based convolution to accelerate the GI loss evaluation, our approach substantially reduces GPU memory usage and training time. However, this efficiency relies on a fixed regular grid, which can limit local resolution. To improve local accuracy in strong scattering regions, we also propose a hybrid GI+PDE loss, enforcing a lightweight Helmholtz residual at a small number of nonuniformly sampled collocation points. We evaluate our method on seismic benchmark models characterized by structural contrasts and subwavelength heterogeneity at frequencies up to 20Hz. GI-based training consistently outperforms PDE-based PINNs, reducing computational cost by over a factor of ten. In models with localized scattering, the hybrid loss yields the most accurate reconstructions, providing a stable, efficient, and physically grounded alternative.

cs.LG↗
arXiv2026

SWEEP (Seismic Wave Equation Exploration Platform): A Unified Solver Framework for Differentiable Wave Physics

SWEEP (Seismic Wave Equation Exploration Platform) is a unified and extensible wave equation solver library designed for wavefield modeling and inversion. It supports a wide range of wave propagation engines, including acoustic, elastic, attenuative, VTI, TTI, and their Born approximations, among others. With a built-in support for automatic differentiation, the framework enables seamless implementation of full-waveform inversion (FWI), least-squares reverse time migration (LSRTM), and other gradient-based optimization methods. It also features a plug-and-play architecture, allowing easy integration and flexible combination of custom loss functions, multi-GPU computation, neural networks, and more. This makes Sweep a powerful and customizable platform for tackling advanced seismic inverse problems.

physics.gen-ph↗
arXiv2026

LAViG-FLOW: Latent Autoregressive Video Generation for Fluid Flow Simulations

Modeling and forecasting subsurface multiphase fluid flow fields underpin applications ranging from geological CO2 sequestration (GCS) operations to geothermal production. This is essential for ensuring both operational performance and long-term safety. While high fidelity multiphase simulators are widely used for this purpose, they become prohibitively expensive once many forward runs are required for inversion purposes and to quantify uncertainty. To tackle this challenge, we propose LAViG-FLOW, a latent autoregressive video generation diffusion framework that explicitly learns the coupled evolution of saturation and pressure fields. Each state variable is compressed by a dedicated 2D autoencoder, and a Video Diffusion Transformer (VDiT) models their coupled distribution across time. We first train the model on a given time horizon to learn their coupled relationship and then fine-tune it autoregressively so it can extrapolate beyond the observed time window. Evaluated on an open-source CO2 sequestration dataset, LAViG-FLOW generates saturation and pressure fields that stay consistent across time while running two orders of magnitude faster than traditional numerical solvers.

cs.LG↗
arXiv2026

Target noise: A pre-training based neural network initialization for efficient high resolution learning

Weight initialization plays a crucial role in the optimization behavior and convergence efficiency of neural networks. Most existing initialization methods, such as Xavier and Kaiming initializations, rely on random sampling and do not exploit information from the optimization process itself. We propose a simple, yet effective, initialization strategy based on self-supervised pre-training using random noise as the target. Instead of directly training the network from random weights, we first pre-train it to fit random noise, which leads to a structured and non-random parameter configuration. We show that this noise-driven pre-training significantly improves convergence speed in subsequent tasks, without requiring additional data or changes to the network architecture. The proposed method is particularly effective for implicit neural representations (INRs) and Deep Image Prior (DIP)-style networks, which are known to exhibit a strong low-frequency bias during optimization. After noise-based pre-training, the network is able to capture high-frequency components much earlier in training, leading to faster and more stable convergence. Although random noise contains no semantic information, it serves as an effective self-supervised signal (considering its white spectrum nature) for shaping the initialization of neural networks. Overall, this work demonstrates that noise-based pre-training offers a lightweight and general alternative to traditional random initialization, enabling more efficient optimization of deep neural networks.

cs.LG↗
arXiv2026

Propagating the prior from far to near offset: A self-supervised diffusion framework for progressively recovering near-offsets of towed-streamer data

In marine towed-streamer seismic acquisition, the nearest hydrophone is often two hundred meter away from the source resulting in missing near-offset traces, which degrades critical processing workflows such as surface-related multiple elimination, velocity analysis, and full-waveform inversion. Existing reconstruction methods, like transform-domain interpolation, often produce kinematic inconsistencies and amplitude distortions, while supervised deep learning approaches require complete ground-truth near-offset data that are unavailable in realistic acquisition scenarios. To address these limitations, we propose a self-supervised diffusion-based framework that reconstructs missing near-offset traces without requiring near-offset reference data. Our method leverages overlapping patch extraction with single-trace shifts from the available far-offset section to train a conditional diffusion model, which learns offset-dependent statistical patterns governing event curvature, amplitude variation, and wavelet characteristics. At inference, we perform trace-by-trace recursive extrapolation from the nearest recorded offset toward zero offset, progressively propagating learned prior information from far to near offsets. The generative formulation further provides uncertainty estimates via ensemble sampling, quantifying prediction confidence where validation data are absent. Controlled validation experiments on synthetic and field datasets show substantial performance gains over conventional parabolic Radon transform baselines. Operational deployment on actual near-offset gaps demonstrates practical viability where ground-truth validation is impossible. Notably, the reconstructed waveforms preserve realistic amplitude-versus-offset trends despite training exclusively on far-offset observations, and uncertainty maps accurately identify challenging extrapolation regions.

physics.geo-ph↗
arXiv2025

Diffusion priors enhanced velocity model building from time-lag images using a neural operator

Velocity model building serves as a crucial component for achieving high precision subsurface imaging. However, conventional velocity model building methods are often computationally expensive and time consuming. In recent years, with the rapid advancement of deep learning, particularly the success of generative models and neural operators, deep learning based approaches that integrate data and their statistics have attracted increasing attention in addressing the limitations of traditional methods. In this study, we propose a novel framework that combines generative models with neural operators to obtain high resolution velocity models efficiently. Within this workflow, the neural operator functions as a forward mapping operator to rapidly generate time lag reverse time migration (RTM) extended images from the true and migration velocity models. In this framework, the neural operator is acting as a surrogate for modeling followed by migration, which uses the true and migration velocities, respectively. The trained neural operator is then employed, through automatic differentiation, to gradually update the migration velocity placed in the true velocity input channel with high resolution components so that the output of the network matches the time lag images of observed data obtained using the migration velocity. By embedding a generative model, trained on a high-resolution velocity model distribution, which corresponds to the true velocity model distribution used to train the neural operator, as a regularizer, the resulting predictions are cleaner with higher resolution information. Both synthetic and field data experiments demonstrate the effectiveness of the proposed generative neural operator based velocity model building approach.

cs.LG↗
arXiv2025

Diffusion Model-Based Posterior Sampling in Full Waveform Inversion

Bayesian full waveform inversion (FWI) offers uncertainty-aware subsurface models; however, posterior sampling directly on observed seismic shot records is rarely practical at the field scale because each sample requires numerous wave-equation solves. We aim to make such sampling feasible for large surveys while preserving calibration, that is, high uncertainty in less illuminated areas. Our approach couples diffusion-based posterior sampling with simultaneous-source FWI data. At each diffusion noise level, a network predicts a clean velocity model. We then apply a stochastic refinement step in model space using Langevin dynamics under the wave-equation likelihood and reintroduce noise to decouple successive levels before proceeding. Simultaneous-source batches reduce forward and adjoint solves approximately in proportion to the supergather size, while an unconditional diffusion prior trained on velocity patches and volumes helps suppress source-related numerical artefacts. We evaluate the method on three 2D synthetic datasets (SEG/EAGE Overthrust, SEG/EAGE Salt, SEAM Arid), a 2D field line, and a 3D upscaling study. Relative to a particle-based variational baseline, namely Stein variational gradient descent without a learned prior and with single-source (non-simultaneous-source) FWI, our sampler achieves lower model error and better data fit at a substantially reduced computational cost. By aligning encoded-shot likelihoods with diffusion-based sampling and exploiting straightforward parallelization over samples and source batches, the method provides a practical path to calibrated posterior inference on observed shot records that scales to large 2D and 3D problems.

physics.geo-ph↗
arXiv2025

Physics-informed waveform inversion using pretrained wavefield neural operators

Full waveform inversion (FWI) is crucial for reconstructing high-resolution subsurface models, but it is often hindered, considering the limited data, by its null space resulting in low-resolution models, and more importantly, by its computational cost, especially if needed for real-time applications. Recent attempts to accelerate FWI using learned wavefield neural operators have shown promise in efficiency and differentiability, but typically suffer from noisy and unstable inversion performance. To address these limitations, we introduce a novel physics-informed FWI framework to enhance the inversion in accuracy while maintaining the efficiency of neural operator-based FWI. Instead of relying only on the L2 norm objective function via automatic differentiation, resulting in noisy model reconstruction, we integrate a physics constraint term in the loss function of FWI, improving the quality of the inverted velocity models. Specifically, starting with an initial model to simulate wavefields and then evaluating the loss over how much the resulting wavefield obeys the physical laws (wave equation) and matches the recorded data, we achieve a reduction in noise and artifacts. Numerical experiments using the OpenFWI and Overthrust models demonstrate our method's superior performance, offering cleaner and more accurate subsurface velocity than vanilla approaches. Considering the efficiency of the approach compared to FWI, this advancement represents a significant step forward in the practical application of FWI for real-time subsurface monitoring.

physics.geo-ph↗
arXiv2025

High-resolution velocity model estimation with neural operator and the time-shift imaging condition

Extracting subsurface velocity information from seismic data is mainly an undetermined problem that requires injecting a priori information to constrain the inversion process. Machine learning has offered a platform to do so through the training process, as we formulate our training dataset to inject as much prior knowledge as possible in the trained ML model. Here, we use a neural-operator-based framework for high-resolution seismic velocity model building, which integrates Neural Operators with time-lag reverse time migration imaging. Unlike conventional full waveform inversion (FWI) methods that rely on iterative forward and adjoint-state computations, our approach learns a direct mapping from initial velocity models and extended seismic images to high-resolution velocity estimates through supervised learning. The network architecture enables mesh-independent generalization by learning mappings in infinite-dimensional function spaces, while synthetic velocity models and time-lag reverse time migration (RTM) images provide complementary high-frequency information critical for recovering mid- and high-wavenumber velocity components. Synthetic experiments demonstrate that the proposed method accurately reconstructs fine-scale structures and complex geologies, including out-of-distribution features such as salt bodies. Applications to real seismic field data acquired offshore Australia further validate the method's robustness and resolution capability. The predicted models show enhanced structural details and improved consistency with well-log data, outperforming traditional multi-scale FWI in both accuracy and computational efficiency. The entire prediction process is completed within seconds, making the proposed approach highly suitable for rapid and scalable velocity model building in practical exploration scenarios.

physics.geo-ph↗
arXiv2025

Velocity model building from seismic images using a Convolutional Neural Operator

The success of building a high-resolution velocity model using machine learning is hampered by generalization limitations that often limit the success of the approach on field data. This is especially true when relying on neural operators for the mapping. Thus, we propose a novel inversion framework that relies on learning to map the velocity model to a seismic image using a Convolutional Neural Operator (CNO), and then we use optimization to invert for the velocity that matches the image. The key to the success of our network is that we use the initial and true velocity models as input in the training, then we invert for the true velocity starting from the initial velocity at inference. Specifically, we first train a neural operator to accurately learn the forward mapping from seismic velocity models to RTM images, using synthetic datasets that include high-frequency structural information. Once trained, the neural operator is embedded into an inversion loop, where its differentiable nature enables efficient gradient computation via automatic differentiation. This allows us to progressively inject high-wavenumber information from RTM images into the background velocity model, thereby improving resolution without the need for traditional adjoint-state solvers. The proposed framework is validated on both synthetic and field data. Results demonstrate that the neural operator generalizes well to real seismic scenarios, maintains high inversion accuracy, and significantly reduces computational cost. This work highlights the potential of neural operators as flexible and scalable tools for efficient, data-driven seismic imaging and inversion.

physics.geo-ph↗