Shearwater Health

SUMMARY Ocean-bottom seismic acquisitions are gaining widespread popularity across a variety of subsurface applications. However, the high cost of these systems often necessitates receiver geometries with large intervals between ocean-bottom cables or nodes. The upside-down Rayleigh–Marchenko (UD–RM) method has been recently proposed as an effective solution for accurate redatuming and imaging of sparse seabed data. In this paper, we present the first successful application of the UD–RM method to field data. We demonstrate that in the presence of a shallow seabed, an improved data pre-processing workflow is necessary to generate more accurate input wavefields compared to the one produced by the workflow presented in the original paper. To validate the proposed processing workflow, the UD–RM method is initially tested on a synthetic data set that mimics the Volve field data (referred to as the Volve synthetic data set); this is followed by its application to a 2-D line of the Volve ocean-bottom cable data set. Subsequently, the field data set is subsampled by retaining only 25 per cent of the total receivers to numerically validate the UD–RM method’s capability to handle sparse receiver arrays. The resulting images reveal that the UD–RM method, when paired with our enhanced data processing workflow, can effectively handle surface-related multiples, internal multiples, and sparse receiver arrays, producing accurate imaging results without the need for costly and labour-intensive multiple removal processes. Finally, we provide theoretical insights and numerical evidence supporting the necessity of source-side deghosting prior to redatuming. While a pre-processing workflow that omits source-side deghosting can offer some practical advantages, we show that this ultimately produces blurrier images compared to those obtained using source-side deghosted input data.

We present the acquisition details and the result of processing and imaging of a marine vibrator (MV) alphatest, which was recently successfully completed over the permanent monitoring system of the Johan Sverdrup field in the North Sea. The alpha-test includes the acquisition of a 3D swath of data using a single MV, several test lines using a two-unit source array covering 3-150Hz, and two ultra low-frequency single-unit lines covering 1-8Hz up to 18km offset. The preliminary result of processing and imaging of the3D swath produced images comparable to the legacy airgun array acquisition. Furthermore, our ultra-low frequency data show valuable signals even at 1.5Hz. Our results clearly show the realistic possibility of employing the MV for the acquisition of broadband seismic data at large scales.

Radon transformation is a popular tool in seismic data processing, imaging, and analysis. The Radon domain is used for deconvolution, multiple suppression, noise muting, and interpolation for pre-processing. Radon transformed data are also used for beam and plane-wave imaging. Data with various curvature are separated in the transformed domain which is a powerful tool for seismic processing and muting. Commonly used methods are linear, parabolic, and hyperbolic. Parabolic and hyperbolic methods have proven to be very useful for normal moveout-corrected data where only flat and curved events are present. Linear Radon is useful for deconvolution, de-multiple and noise muting. Seismic data may have different curvatures along different axes; to handle this issue we propose a mixed curvature Radon transformation, which applies linear and parabolic transformations in the X and Y directions separately or vice versa. We show both synthetic and real data examples to demonstrate that our proposed method is a good alternative for input data with mixed curvatures.

Laboratory stick-slip experiments are a simple analogue for the earthquake cycle. The acoustic emissions (AE) of these experiments have been shown to contain hidden patterns. Machine Learning (ML) can extract these patterns and information on the fault state can be inferred (e.g. shear stress and time to failure). Two different ML approaches have been used in the past: 1) ensemble tree models, which are relatively easy to evaluate why they made a certain prediction, but only look at a snapshot in time and 2) deep neural networks using Long Short-Term Memory (LSTM), which have the ability to find patterns in the temporal changes in the signal, but act more as a black-box model, so the final predictions are hard to evaluate. Here we introduce an additional step in the workflow that can be used to allow the ensemble tree models information about the temporal changes of the input features. Furthermore, it is able to quantify and visualize whether a pattern is repetitive or not. Like earlier studies we start by calculating (statistical) features using a rolling window on the AE. The features are not directly used as the input of the model, but are placed in a larger Hankel matrix, where the consecutive time windows are the rows of the matrix. Using Principal Component Analysis (PCA) and Uniform Manifold Approximation and Projection (UMAP) we create an embedded version of this array that holds temporal information of features calculated in the previous step. Visual inspection of these embeddings shows that some features map to very distinct patterns that are repetitive over the majority of the stick-slip cycles. The advantage of this method is that an inverse mapping is easily available, allowing for an interpretable embedding of the data.

Summary Recently, the first-ever 3D broadband survey utilizing towed marine vibrators (MVs) was successfully acquired over a North Sea permanent reservoir monitoring (PRM) system. Two vibrators were deployed, one low-band unit and one high-band unit, collectively covering the entire seismic frequency band from 3 to 150 Hz. The implementation of phase encoding proved effective in efficiently reducing residual sweep noise. Consequently, a nearly 100% utilization rate was achieved, eliminating the need for a silent period between sweeps. In an additional test, the survey demonstrated the emission of robust signals in the 1–8 Hz range, which holds significance for Full Waveform Inversion (FWI) applications and imaging. The survey encompassed comprehensive testing and verification, including towing and handling, positioning, control software assessment, source signature measurements, and Quality Control (QC) measures, forming a holistic acquisition system. In comparison to airgun-based sources, marine vibrators exhibited a notable decrease in peak pressure and out-of-band (high-frequency) noise pollution. The processed MV data displayed favorable comparisons with legacy airgun-based data, highlighting the realistic potential of employing marine vibrators for large-scale acquisition of broadband seismic data.

In multicomponent ocean bottom seismic (OBS) acquisition, hydrophones and geophones (or accelerometers) are employed to record seismic signals. Combining records from the hydrophone (P-component) and the vertical geophone (Z-component) allows for elimination of the receiver-side ghost and water-layer multiples, when performing wavefield separation above and below the seafloor respectively. Apart from the environmental noise that affects both hydrophone and geophone records, the Zcomponent is often contaminated by an additional type of shear-like noise known as Vz noise. Unless removed, Vz noise can cause significant issues with subsequent processing steps. There have been several works on attenuating Vz noise with different levels of success. Here we propose an effective deep learning (DL) approach for this task.

Marine seismic vibrators are considered serious contenders to air-gun arrays and are attractive as environmentally friendly sources. While the reduced environmental impact is highly attractive, an additional driver is the ability to control the characteristics of the released energy precisely. Here we show how to acquire and process seismic data with marine vibrators while employing a frequency-dependent spatial sampling up to 150Hz and a water depth of 145m, extending previous works which were up to 50Hz with 1900m water depth. Furthermore, we review the key processing challenges associated with this advanced acquisition design. We also detail and emphasize the estimation of the source signature for accurate and effective sweep deconvolution.

Full-waveform inversion (FWI) aims to obtain accurate subsurface models by minimizing the discrepancy between observed seismic data and modeled data. However, the commonly used L2-norm waveform difference misfit functional is prone to cycle-skipping due to local minima. To address this issue, we propose more effective misfit metrics for FWI by utilizing the soft-dynamic time warping (SDTW) divergence distance and its sharp variant. The proposed methods introduce a hyper-parameter to ensure differentiability of the functional, enabling the use of the adjoint state method for gradient computation in FWI. Unlike conventional SDTW, our divergence-based metrics always yield positive values, reaching their minimum when the modeled trace matches the observed trace. The efficacy of the proposed methods is demonstrated through their application on a field dataset, highlighting their robustness in mitigating cycle-skipping compared to the conventional L2 norm.