Mehdi Aharchaou's research while affiliated with ExxonMobil and other places

The high resolving power of seismic measurements has promoted wide adoption of the seismic method in oil and gas and other industries. Studying the evolution of seismic resolution, the different factors affecting it, and the remaining barriers enables an improved understanding of where we are today and what lies ahead. The need to improve seismic resolution is best framed in the context of the interpretation questions being raised and the project stage (e.g., new frontier, appraisal, development, or production). Improvements in resolution do not depend on a single aspect of the seismic workflow but on multiple interconnected components including acquisition, processing, imaging, and interpretation methods and technologies. This paper highlights some of the key milestones in improving seismic resolution. We also conjecture on progress likely to be made in the years ahead and remaining opportunities to enhance seismic resolution.

The science of modern seismology was born more than 100 years ago (1889) when the first teleseismic record was identified and the seismograph was developed ( Ben-Menahem, 1995 ). In 1921, earth exploration was revolutionized when a team led by Clarence Karcher conducted the first field tests of the reflection seismograph in Oklahoma City ( Dragoset, 2005 ). That experiment showed that the subsurface can be imaged using seismic data. Businesses boomed as the seismic method started establishing its track record in finding hydrocarbons. Over the last century, the seismic method has emerged as the cornerstone of exploration geophysics, providing us with increasingly accurate characterizations of the subsurface and enabling us to better discover and describe hydrocarbon prospects, geothermal anomalies, seafloor hazards, aquifers, and much more.

We propose a workflow to correct migration amplitudes by estimating the inverse Hessian operator weights using a neural network based framework. We train the network such that it learns the transformation between the migration output and true amplitude reflectivity constrained by different conditioners. We analyze the network output with velocity model and with source illumination as a conditioner. As compared to the velocity model, source illumination as a conditioner performs better because source illumination encodes the geometrical spreading information and accounts for non‐stationarity. We further use the output of the deep neural network as a starting model for accelerating the convergence of iterative least‐squares reverse time migration. Using a deep learning framework, the proposed method combines the model domain and data domain least‐squares migration approaches to recover images with interpretable amplitudes, attenuated migration artifacts, better signal‐to‐noise ratio, and improved resolution. We compare the output of the proposed algorithm with conventional least‐squares output and show that the proposed workflow is more robust, especially in the areas with weak illumination. This article is protected by copyright. All rights reserved

Shallow gas accumulations with varied spatial distribution and scales cause amplitude loss and phase distortion in seismic images in underlying reservoir intervals. We present an integrated velocity and attenuation (defined by the quality factor Q) model-building workflow to address the imaging challenges in a shallow-water data set acquired offshore Sakhalin Island, Russia. As part of the workflow, we use full-waveform inversion (FWI) to capture fine details of the velocity and Q models. We achieve a step change in image quality improvement and amplitude fidelity enhancement by leveraging the workflow. This image enhancement improves subsurface interpretation and influences well placement, subsurface model building, and depletion plan optimization. Key to this success is not only the advanced model-building tool kit, which includes FWI and tomography, but also the optimal integration of the tools specifically tailored for the geologic setting and imaging challenges offshore Sakhalin Island.

Recent demands to reduce turnaround times and expedite investment decisions in seismic exploration have invited new ways to process and interpret seismic data. Among these ways is a more integrated collaboration between seismic processors and geologist interpreters aiming to build preliminary geologic models for early business impact. A key aspect has been quick and streamlined delivery of clean high-fidelity 3D seismic images via postmigration filtering capabilities. We present a machine learning-based example of such a capability built on recent advances in deep learning systems. In particular, we leverage the power of Siamese neural networks, a new class of neural networks that is powerful at learning discriminative features. Our novel adaptation, edge-aware filtering, employs a deep Siamese network that ranks similarity between seismic image patches. Once the network is trained, we capitalize on the learned features and self-similarity property of seismic images to achieve within-image stacking power endowed with edge awareness. The method generalizes well to new data sets due to the few-shot learning ability of Siamese networks. Furthermore, the learning-based framework can be extended to a variety of noise types in 3D seismic data. Using a convolutional architecture, we demonstrate on three field data sets that the learned representations lead to superior filtering performance compared to structure-oriented filtering. We examine both filtering quality and ease of application in our analysis. Then, we discuss the potential of edge-aware filtering as a data conditioning tool for rapid structural interpretation.

The lack of seismic low frequencies in towed-streamer data is known to have an outsized detrimental effect on advanced velocity model building techniques such as full-waveform inversion (FWI). Since seabed acquisition records ultralow frequencies (1–4 Hz) with high signal-to-noise ratio, this presents an opportunity to learn, in a supervised machine learning fashion, a bandwidth extension function to enrich towed-streamer data with low frequencies. We use recent advances in training deep neural networks to develop a novel method for learning low-frequency reconstruction from an ultrasparse set of ocean-bottom nodes (OBNs). This bandwidth extension is tested on two large field data sets (from an OBN survey and a wide-azimuth towed-steamer survey) acquired over a complex-shaped salt region in the Gulf of Mexico. The reconstructed low frequencies, although not perfect, enable FWI to more effectively correct the shape of salt bodies and result in improved subsalt imaging. Well-tie analysis shows an improvement in phase stability around the wellbore and a fit to within half a cycle at reservoir level. This work links together towed-streamer and seabed acquisitions, providing a cost-effective solution to help offshore seismic exploration with higher-quality low frequencies.

Broadband preprocessing has become widely employed for marine towed-streamer seismic data. In the standard workflow, farfield source designature, receiver and source-side deghosting, and redatuming to mean-sea-level are applied in sequence, with amplitude compensation for background Q delayed until the imaging or post migration stages. Thus each step is likely to generate its own artifacts, quality checking can be time consuming, and broadband data are only obtained late in this chained workflow. We present a unified method for broadband preprocessing—termed integrated broadband preprocessing (IBP) – that enables the joint application of all the above listed steps early in the processing sequence. Amplitude, phase and AVO fidelity of IBP are demonstrated on pressure data from shallow, deep and slanted streamers. The full integration allows greater sparsity to emerge in the representation of seismic data, conferring clear benefits over the sequential application. Moreover, time sparsity, full dimensionality and early amplitude Q compensation all have an impact on broadband data quality, in terms of reduced ringing artifacts, improved wavelet integrity at large crossline angles and fewer residual high-frequency multiples.