-
Poststack Seismic Data Preconditioning via Dynamic Guided Learning
Authors:
Javier Torres-Quintero,
Paul Goyes-Peñafiel,
Ana Mantilla-Dulcey,
Luis Rodríguez-López,
José Sanabria-Gómez,
Henry Arguello
Abstract:
Seismic data preconditioning is essential for subsurface interpretation. It enhances signal quality while attenuating noise, improving the accuracy of geophysical tasks that would otherwise be biased by noise. Although classical poststack seismic data enhancement methods can effectively reduce noise, they rely on predefined statistical distributions, which often fail to capture the complexity of s…
▽ More
Seismic data preconditioning is essential for subsurface interpretation. It enhances signal quality while attenuating noise, improving the accuracy of geophysical tasks that would otherwise be biased by noise. Although classical poststack seismic data enhancement methods can effectively reduce noise, they rely on predefined statistical distributions, which often fail to capture the complexity of seismic noise. On the other hand, deep learning methods offer an alternative but require large and diverse data sets. Typically, static databases are used for training, introducing domain bias, and limiting adaptability to new noise poststack patterns. This work proposes a novel two-process dynamic training method to overcome these limitations. Our method uses a dynamic database that continuously generates clean and noisy patches during training to guide the learning of a supervised enhancement network. This dynamic-guided learning workflow significantly improves generalization by introducing variability into the training data. In addition, we employ a domain adaptation via a neural style transfer strategy to address the potential challenge of encountering unknown noise domains caused by specific geological configurations. Experimental results demonstrate that our method outperforms state-of-the-art solutions on both synthetic and field data, within and outside the training domain, eliminating reliance on known statistical distributions and enhancing adaptability across diverse data sets of poststack data.
△ Less
Submitted 2 February, 2025;
originally announced February 2025.
-
Physically Guided Deep Unsupervised Inversion for 1D Magnetotelluric Models
Authors:
Paul Goyes-Peñafiel,
Umair bin Waheed,
Henry Arguello
Abstract:
The global demand for unconventional energy sources such as geothermal energy and white hydrogen requires new exploration techniques for precise subsurface structure characterization and potential reservoir identification. The Magnetotelluric (MT) method is crucial for these tasks, providing critical information on the distribution of subsurface electrical resistivity at depths ranging from hundre…
▽ More
The global demand for unconventional energy sources such as geothermal energy and white hydrogen requires new exploration techniques for precise subsurface structure characterization and potential reservoir identification. The Magnetotelluric (MT) method is crucial for these tasks, providing critical information on the distribution of subsurface electrical resistivity at depths ranging from hundreds to thousands of meters. However, traditional iterative algorithm-based inversion methods require the adjustment of multiple parameters, demanding time-consuming and exhaustive tuning processes to achieve proper cost function minimization. Recent advances have incorporated deep learning algorithms for MT inversion, primarily based on supervised learning, and large labeled datasets are needed for training. This work utilizes TensorFlow operations to create a differentiable forward MT operator, leveraging its automatic differentiation capability. Moreover, instead of solving for the subsurface model directly, as classical algorithms perform, this paper presents a new deep unsupervised inversion algorithm guided by physics to estimate 1D MT models. Instead of using datasets with the observed data and their respective model as labels during training, our method employs a differentiable modeling operator that physically guides the cost function minimization, making the proposed method solely dependent on observed data. Therefore, the optimization algorithm updates the network weights to minimize the data misfit. We test the proposed method with field and synthetic data at different acquisition frequencies, demonstrating that the resistivity models obtained are more accurate than those calculated using other techniques.
△ Less
Submitted 13 January, 2025; v1 submitted 20 October, 2024;
originally announced October 2024.
-
CDDIP: Constrained Diffusion-Driven Deep Image Prior for Seismic Image Reconstruction
Authors:
Paul Goyes-Peñafiel,
Ulugbek Kamilov,
Henry Arguello
Abstract:
Seismic data frequently exhibits missing traces, substantially affecting subsequent seismic processing and interpretation. Deep learning-based approaches have demonstrated significant advancements in reconstructing irregularly missing seismic data through supervised and unsupervised methods. Nonetheless, substantial challenges remain, such as generalization capacity and computation time cost durin…
▽ More
Seismic data frequently exhibits missing traces, substantially affecting subsequent seismic processing and interpretation. Deep learning-based approaches have demonstrated significant advancements in reconstructing irregularly missing seismic data through supervised and unsupervised methods. Nonetheless, substantial challenges remain, such as generalization capacity and computation time cost during the inference. Our work introduces a reconstruction method that uses a pre-trained generative diffusion model for image synthesis and incorporates Deep Image Prior to enforce data consistency when reconstructing missing traces in seismic data. The proposed method has demonstrated strong robustness and high reconstruction capability of post-stack and pre-stack data with different levels of structural complexity, even in field and synthetic scenarios where test data were outside the training domain. This indicates that our method can handle the high geological variability of different exploration targets. Additionally, compared to other state-of-the-art seismic reconstruction methods using diffusion models. During inference, our approach reduces the number of sampling timesteps by up to 4x.
△ Less
Submitted 14 January, 2025; v1 submitted 24 July, 2024;
originally announced July 2024.
-
GAN-supervised Seismic Data Reconstruction: An Enhanced-Learning for Improved Generalization
Authors:
Paul Goyes-Penafiel,
Leon Suarez-Rodriguez,
Claudia Correa,
Henry Arguello
Abstract:
Seismic data interpolation of irregularly missing traces plays a crucial role in subsurface imaging, enabling accurate analysis and interpretation throughout the seismic processing workflow. Despite the widespread exploration of deep supervised learning methods for seismic data reconstruction, several challenges still remain open. Particularly, the requirement of extensive training data and poor d…
▽ More
Seismic data interpolation of irregularly missing traces plays a crucial role in subsurface imaging, enabling accurate analysis and interpretation throughout the seismic processing workflow. Despite the widespread exploration of deep supervised learning methods for seismic data reconstruction, several challenges still remain open. Particularly, the requirement of extensive training data and poor domain generalization due to the seismic survey's variability poses significant issues. To overcome these limitations, this paper introduces a deep-learning-based seismic data reconstruction approach that leverages data redundancy. This method involves a two-stage training process. First, an adversarial generative network (GAN) is trained using synthetic seismic data, enabling the extraction and learning of their primary and local seismic characteristics. Second, a reconstruction network is trained with synthetic data generated by the GAN, which dynamically adjusts the noise and distortion level at each epoch to promote feature diversity. This approach enhances the generalization capabilities of the reconstruction network by allowing control over the generation of seismic patterns from the latent space of the GAN, thereby reducing the dependency on large seismic databases. Experimental results on field and synthetic seismic datasets both pre-stack and post-stack show that the proposed method outperforms the baseline supervised learning and unsupervised approaches such as deep seismic prior and internal learning, by up to 8 dB of PSNR.
△ Less
Submitted 27 March, 2024; v1 submitted 17 November, 2023;
originally announced November 2023.
