Search

Several methods have been proposed to improve the signal-to-noise ratio by attenuating incoherent noise, including prediction error filtering (Canales 1984), projection filtering (Soubaras 1995), and more recently rank reduction filtering. In this last category, we can differentiate eigenimage filtering (Trickett 2003), Cadzow / Singular Spectrum Analysis (SSA) filtering (Trickett 2009, Sacchi 2009) and tensor methods (Kreimer and Sacchi 2012, Trickett 2013, Da Silva and Herrmann 2014). Also, these latter methods have been extended to robust noise attenuation to deal with erratic noise, or data interpolation for binned data within a defined grid (Trickett et al. 2010, 2012, Oropeza and Sacchi 2011, Chen and Sacchi 2013). Here, we propose a systematic formulation of the simultaneous random plus erratic noise attenuation and data interpolation problem as a convex optimization program, which can be solved efficiently. We model the coherent signal via its low-rank trajectory matrix in the spirit of Cadzow/SSA filtering, and the erratic noise as a sparse component of the input data. The signal component is recovered by solving a Joint Low-Rank and Sparse Inversion (JLRSI) thanks to a joint minimization of the nuclear and L1 norms of the low-rank and sparse components respectively.

One way to address the problem of weak subsalt illumination is through angle gather illumination weighting (AGILW). In this technique, synthetic data mimicking the field data are generated and migrated the same way as the field data. Illumination weighting scalars are obtained by measuring coherent amplitude on the synthetic migration gather. This weighting scalar can be applied to field data to enhance signal and attenuate noise. In this paper, we propose a refinement of this approach: shot patch-based angle gather illumination weighting, based on RTM 3D angle gathers. Instead of using the traditional approach of migrating all shots in a survey to form one set of angle gathers, we partition the input shots from both field and synthetic data into smaller shot patches and keep the migration gathers separate for each shot patch. This new approach can limit crosstalk between signal and noise from different shot patches, and can more effectively attenuate noise and improve subsalt images. We demonstrate the effectiveness of our method with both 2D synthetic data and 3D field data.

The test showed that, in the central North Sea, with appropriate denoising, we were able to successfully remove SI from all azimuths down to a vessel separation of less than 25 km. By adjusting vessel speeds to make sure that SI did not arrive at the same time from shot to shot, even better denoising results were achieved.

We present an input data selection workflow based on 3D ray-tracing to improve the RTM image in areas of poor illumination and low signal-to-noise ratio. It is effective for imaging subsalt three-way closure with weak subsalt primaries and strong noise levels. The workflow can be applied on any type of survey, but it is most suitable for full azimuth geometries. We focus on data selection using 3D ray-tracing, but this workflow can be easily adapted to use finite-difference wave-equation modelling. The data selection information can either be used to scale up the weak primary signal before migration or to be migrated separately and merged into a full migration result in the post-migration stage.

Serial processing of PP and PS waves leads to depth inconsistencies between the PP and PS migrated images. The link between P- and S-velocity model is the Vp/Vs ratio, whose estimation is an important task in multicomponent analysis because of its applications to lithology and fluid discrimination. We show the advantages of using the Vp/Vs ratio estimated with dynamic image warping as a volumetric constraint for nonlinear slope tomography. Our approach does not require repeated shift estimation between PP and PS events during inversion and does not depend on picked horizons thus reducing human interaction and uncertainty associated with interpretation.d PS events during inversion and does not depend on picked horizons thus reducing human interaction and uncertainty associated with interpretation.

Discusses use of simultaneous-source acquisition for sparse (ROV-deployed) OBN surveys, using time and motion studies to quantify impact on acquisition time. Also describes new technique for sim-src crosstalk noise attenuation using modified Radon operators. Finally, application to Gorgon OBN data for baseline and monitor, with PSDM images and 4D differences. Concludes with a 10 % time saving for sim-src on the Gorgon OBN acquisition, and an additional NRMS penalty on the 4D difference of 6 % after crosstalk attenuation. Without special processing, the crosstalk adds 70 % NRMS to the final images.

Multiple timelapse realisations produced using wavefield separation, data selection, or diverging processing flows are combined using powerful noise-reducing data weights formulated using the 3D images as well as the 4D differences. The result separates 4D signal from incoherent 4D noise and also 4D noise that is coherent across multiple realisations of 4D difference but not repeated in both baseline and monitor 3D images (e.g. residual multiple in the monitor). Examples using Broadseis base and monitor from the North Sea, and towed-streamer base and OBN monitor from deep water offshore Angola, show considerable uplift to the final 4D difference.

In the context of WAZ data the VTI hypothesis is not always sufficient for insuring focusing of time migration. We propose an extension of non-linear slope tomography for time imaging to the orthorhombic case. We use a model of orthorhombic anisotropy parameterized by five effective parameters and where the effective velocity and eta vary according to the azimuth of the migrated trace. In our approach the five parameters are updated jointly, allowing extending to the orthorhombic case the advantages of non-linear slope tomography for time velocity model building, i.e. and improved accuracy and turn-around.

The seismic image represents the spatially variable reflectivity of the medium where migration effectively rotates the wavelet to be normal to the imaged reflectors. While this is the general case, it is often disregarded, and one-dimensional spectral analysis of the vertical coordinate is commonly used. We show that spectral processing of post-imaging data in the direction normal to the reflectors provides accurate results for steep and complex structures. We introduce the concept of Orthogonal-Image-Gathers (OIGs) which facilitate this approach while providing a platform for handling spatially variable spectral distortions due to the velocity field and other medium-related properties.