Remco Muijs

Prestack depth migration of primary and surface-related multiple reflections: Part I — Imaging

Surface-related multiples (i.e., all seismic waves reflected at the free surface at least once) often severely contaminate seismic recordings. Because conventional imaging techniques require input data that consist of primary reflections only, significant processing effort is commonly dedicated to attenuating multiples prior to migration. On the other hand, surface-related multiples provide additional illumination of the subsurface and, therefore, should not be considered as noise. We present a prestack depth-migration method that allows primary and multiple reflections to be imaged simultaneously. Depth imaging using primary and multiple reflections (DIPMR) involves decomposing the datainto upgoing and downgoing wave constituents, followed by downward extrapolation. Artifacts generated by interference of upgoing and downgoing events not associated with the same subsurface reflection points (crosstalk) are attenuated by using a 2D deconvolution imaging condition. In contrast to existing methods, DIPMR doe...

Prestack depth migration of primary and surface‐related multiple reflections

Surface-related multiples may severely contaminate multicomponent seabed seismic recordings. Because conventional imaging techniques require input data that consist of primary reflections only, significant processing effort is commonly dedicated to attenuating multiples prior to migration. On the other hand, surface-related multiples provide additional illuminations of the subsurface. We present a prestack depthmigration method that allows primary and multiple reflections to be imaged simultaneously. Depth imaging with primary and multiple reflections (DIPMR) involves decomposing the data into upand downgoing wave constituents, followed by downward extrapolation. Artifacts generated by interference of upand downgoing events not associated with the same subsurface reflection points (cross-talk) are attenuated using a 2D deconvolution imaging condition. In contrast to existing methods, DIPMR does not require a priori information about the source signature or directivity, because the illuminating source wavefield is extracted directly from the data themselves via the up/down separation. Moreover, there is no need for elimination or identification of multiples prior to migration. By including surface-related multiples in the imaging procedure, the effective source wavefield is stronger and the spatial aperture wider.

Prestack depth migration of primary and surface-related multiple reflections: Part II — Identification and removal of residual multiples

Depth imaging using primary and multiple reflections (DIPMR), as described in Part I of this study, allows subsurface information carried by multiple reflections to be utilized. In the presence of strong lateral heterogeneity, however, the migration results may be distorted by artifacts originating from reflections associated with layers above the imaging plane that are commonly referred to as crosstalk. We present an image enhancement procedure that allows such artifacts to be effectively suppressed by predicting the initial crosstalk in a second migration phase. This second migration uses reflections imaged at shallower depth levels and requires knowledge of the total and primary downgoing wavefield at the receiver level. The predicted crosstalk image it-self may assist the interpretational effort by indicating areas where artifacts may result in incorrect identification of geologic structures or may cause local distortions of amplitude informa-tion. Furthermore, a clean depth image can be obtained by a...

Data‐driven adaptive decomposition of multicomponent seabed recordings

Dual-sensor (hydrophone and three-component geophone) data recorded on the seafloor allow the elastic wavefield to be decomposed into its upgoing and downgoing P- and S-wave components. Most decomposition algorithms require accurate knowledge of the elastic properties of the seafloor in the vicinity of the receivers and properly calibrated sensors, in order for the data to be a faithful vector representation of the ground motion. We present a multistep adaptive decomposition scheme that provides the necessary information directly from the data by imposing constraints on intermediate decomposition results. The proposed scheme requires no a priori information and only a minimal amount of user-defined input, thus allowing multicomponent data to be decomposed in an automated data-driven fashion. The performance of the technique is illustrated using seabed data acquired in the North Sea with prototype single sensors (multicomponent geophones individually sampled). Realistic seafloor properties and sensor calibration operators are obtained, and elastic decomposition of the calibrated data generally yields good results. Dominant water-layer reverberations are successfully attenuated and primary reflections are substantially enhanced in the computed upgoing P-wave potential just below the seafloor. In contrast, the result for the upgoing S-wave potential is somewhat less convincing; although the energy of water-layer multiples is substantially reduced, notable amounts of undesired multiple energy remain in this section after decomposition, particularly at high offsets. These imperfections may point to inaccuracies in the parametrization of the seafloor or remaining inaccuracies in the vector fidelity of the horizontal geophone recordings. Nevertheless, the results obtained with the extended data-driven decomposition scheme are at least comparable to previously published results.

Perturbation analysis of an explicit wavefield separation scheme for P‐ and S‐waves

Dense spatial recording patterns of three‐component (3C) receivers allow for direct wavefield decomposition through explicit calculation of divergence and curl of the recorded elastic wavefield. Since this approach is based upon the observation of small phase shifts, it requires highly accurate deployment of the receiver configurations. To study the feasibility of a recently proposed P/S‐wave separation scheme, we systematically assess the effects of position and orientation errors of one or several geophones within the recording pattern on technique performance.We find that realistic deployment errors can significantly affect estimates of the divergence and curl of particle velocity. The errors induced by mispositioned or misoriented geophones differ for each of the geophones that make up a pattern. Moreover, the inaccuracies vary with the angle of incidence, potentially affecting analysis procedures applied to the data at a later stage, such as amplitude variation with offset (AVO). Based on a relative ...

Data-driven adaptive decomposition of multicomponent seabed seismic recordings: Application to shallow-water data from the North Sea

Exploiting the full potential of multicomponent seabed seismic recordings requires the decomposition of the recorded data into their upgoing and downgoing P- and S-wave constituents. We present a case study from the North Sea, where a novel adaptive wave-equation-based decomposition method is applied to a 2D data set shot inline with a cable-based seabed seismic acquisition system. The data were recorded in relatively shallow (∼150 m) water, such that severe interference exists between primary reflections and water-layer multiples. Such conditions represent a challenge for many decomposition methods, because these often require a significant amount of interpretive, user-defined input. Conversely, the adaptive algorithm demonstrated in this study is fully data-driven, requiring as sole input a rough estimate of the water depth. The importance of careful mutual calibration of the sensors is demonstrated by critically assessing the properties of the derived calibration filters and the resulting estimates of ...

Near‐surface seismic properties for elastic wavefield decomposition: Estimates based on multicomponent land and seabed recordings

Accurate knowledge of the seismic material properties in the immediate vicinity of the receivers represents a prerequisite for elastic wavefield decomposition. We present strategies for estimating the elastic material properties for both land and seabed multicomponent seismic data. The proposed scheme for land data requires dense multicomponent geophone configurations, which allow spatial wavefield derivatives to be explicitly calculated. The required information can be obtained with four three-component surface geophones positioned at the corners of a square, and a fifth geophone buried at a shallow depth below the center of the square. The technique yields local estimates of the near-surface P- and S-wave velocities, but the density cannot be constrained. Using a similar approach for four-component (three orthogonal components of particle velocity plus pressure) seabed recordings allows the P- and S-wave velocities as well as the density of the seafloor to be estimated. In this case, the proposed scheme does not require buried geophones, and it is applicable to multicomponent data recorded in routine seabed surveys. Compared to existing techniques, the new method allows the elastic seafloor properties to be more accurately determined, and it does not rely critically on the inclusion of large-offset data. Numerical tests indicate that the proposed schemes are robust and yield accurate results, provided that the signal used for the inversion contains sufficient horizontal energy and can be clearly identified and separated from other signals. Although the schemes are designed for application on the first arrivals, they are, in principle, applicable to any data window containing isolated P- or S-arrivals. The proposed scheme is successfully applied to a seabed data set acquired in the North Sea. In contrast, the application on a multicomponent land data set was unsuccessful, because of strong receiver-to-receiver variations in amplitude and phase, probably caused by differences in coupling and instrument response.

Non-interactive Estimation of Elastic Sea Floor Properties For Wavefield Decomposition

The advent of multi-component seismic recordings at the seabed offers the perspective to decompose the wavefield into its up- and downgoing P- and S-wave components. Most decomposition algorithms require (i) accurate knowledge of the elastic properties of the sea floor in the immediate vicinity of the receivers and (ii) properly calibrated sensors such that the data are a faithful vector representation of the actual ground motion. In this paper, we present a strategy to accurately estimate the necessary information by imposing constraints on inter- mediate decomposition results. The proposed scheme is based on a recent adaptive decomposition technique suggested by Schalkwijk et al. (1999) and shares its feature of being completely data-driven in the sense that it requires no a priori information, such as water depth or source signature. Our scheme, however, is based on the observation that imperfect wave field decomposition results in energy arriving simultaneously in suitable combinations of wavefield constituents.

Estimation of inelastic seismic material properties of a surficial sea-bed from multi-component marine seismic data

To date, state-of-the-art seismic material parameter estimates from multi-component sea-bed seismic data are based on the assumption that the sea-bed consists of a fully elastic half-space. In reality, however, the shallow sea-bed generally consists of soft, unconsolidated sediments that are characterized by strong to very strong seismic attenuation. To explore the potential implications, we apply a state-of-the-art elastic decomposition algorithm to synthetic data for a range of canonical sea-bed models consisting of a viscoelastic half-space of varying attenuation. We find that in the presence of strong seismic attenuation, as quantified by Q -values of 10 or less, significant errors arise in the conventional elastic estimation of seismic properties. Tests on synthetic data indicate that these errors can be largely avoided by accounting for the inherent attenuation of the seafloor when estimating the seismic parameters. This can be achieved by replacing the real-valued expressions for the elastic moduli in the governing equations in the parameter estimation by their complex-valued viscoelastic equivalents. The practical application of our parameter procedure yields realistic estimates of the elastic seismic material properties of the shallow sea-bed, while the corresponding Q -estimates seem to be biased towards too low values, particularly for S-waves. Given that the estimation of inelastic material parameters is notoriously difficult, particularly in the immediate vicinity of the sea-bed, this is expected to be of interest and importance for civil and ocean engineering purposes.

P/vz‐calibration of multicomponent seabed data for wavefield decomposition

Acquiring multicomponent seismic data directly on the seafloor allows for a decomposition of the wavefield into its upand downgoing Pand S-wave components. Existing decomposition techniques critically rely on the assumption that multicomponent data provide a faithful vector representation of the actual ground motion. In practice, the effects of inconsistent receiver coupling and differences in instrument response need to be accounted for prior to decomposition. In this paper, I present a method for calibration of the recordings of the vertical component of particle velocity (vz) against the pressure (P) by imposing adequate constraints on the upand downgoing pressure just above seafloor. Frequency-dependent calibration operators can be designed by requiring the amplitude spectrum of the upgoing pressure just above the seafloor to be equal to the amplitude spectrum of the downgoing pressure just above the seafloor after muting of the direct wave. A major advantage of this approach is that no information about the water depth is required during the calibration phase and the scheme can be applied in the presence of a dipping seafloor. The technique does not rely on explicit classification of events in the recorded data and its application is therefore both efficient and economical. The performance of the calibration scheme is successfully demonstrated on a multicomponent data set.