Coherent Noise Attenuation Research Articles

A robust polynomial principal component analysis for seismic noise attenuation

Random and coherent noise attenuation is a significant aspect of seismic data processing, especially for pre-stack seismic data flattened by normal moveout correction or migration. Signal extraction is widely used for pre-stack seismic noise attenuation. Principle component analysis (PCA), one of the multi-channel filters, is a common tool to extract seismic signals, which can be realized by singular value decomposition (SVD). However, when applying the traditional PCA filter to seismic signal extraction, the result is unsatisfactory with some artifacts when the seismic data is contaminated by random and coherent noise. In order to directly extract the desired signal and fix those artifacts at the same time, we take into consideration the amplitude variation with offset (AVO) property and thus propose a robust polynomial PCA algorithm. In this algorithm, a polynomial constraint is used to optimize the coefficient matrix. In order to simplify this complicated problem, a series of sub-optimal problems are designed and solved iteratively. After that, the random and coherent noise can be effectively attenuated simultaneously. Applications on synthetic and real data sets note that our proposed algorithm can better suppress random and coherent noise and have a better performance on protecting the desired signals, compared with the local polynomial fitting, conventional PCA and a L1-norm based PCA method.

On Relative Precedence of Coherent and Random Noise Attenuation

Noise attenuation has always been a challenging affair for acquisition or processing geophysicist. The quality of final product and ease for application of next main phases of processing such as deconvolution, velocity analysis, residual statics computation, stacking and migration grossly depends on the way and amount, the noise is eliminated while keeping the signal preserved. This study is aimed at analyzing and recommending the relative precedence of attenuation steps for coherent and random noises. It is inferred, using field data examples that source related coherent noise be treated first for attenuation and the random noise next in the realm of seismic data processing steps.

S/N enhancement by means of array simulation for near surface seismic investigations

ABSTRACTWe present a procedure for enhancing the signal‐to‐noise ratio (S/N) of shallow seismic reflection data based on two different steps: 1) an acquisition step that requires the recording of closely spaced common source records with standard source and receiver equipment and 2) a processing step where weighted or un‐weighted source and receiver arrays are simulated on the basis of required needs for source related noise attenuation and depth penetration. The data acquisition can be carried out employing single source‐single geophone recordings, with a standard 24 or 48‐channel equipment. Simple energy sources such as weight drop or sledgehammer are considered. The design and application of the spatial filters in the processing phase is very flexible and can be tailored to the specific needs. In fact, the simulated source and/or receiver arrays can be time and/or space variant and can be weighted to provide the desired responses. Optimal weights can be determined by means of Chebyshev polynomials. Real data examples show the increase in the data quality in terms of better coherent noise attenuation and of enhanced depth penetration.

Application of the Frequency-Wave Number (F-K) and the Radial Trace Transform (RTT) in the Attenuation of Coherent Noise in Onshore Seismic Data

Application of the Frequency-Wave Number (F-K) and the Radial Trace Transform (RTT) in the Attenuation of Coherent Noise in Onshore Seismic Data

Geophysical Signal Parameterization and Filtering Using the fractional Fourier Transform

The fractional Fourier transform (FrFT) domain operations offer an alternative to the conventional Fourier Transform (FT) in signal processing, especially for seismic data. It is shown that subsurface reflection events in a two-dimensional (time-spatial) seismic data behave like linear chirps for each frequency in the frequency-spatial domain. Since the FrFT kernel is a set of linear chirps, frequency-spatial domain seismic data processing is better suited using the FrFT than the FT. An analytical relationship between linear seismic events and FrFT parameters is derived and illustrated with an example. Seismic data from a field survey is used to show that the FrFT filtering performs better in coherent noise attenuation than the FT.

Seismic coherent and random noise attenuation using the undecimated discrete wavelet transform method with WDGA technique

One of the most crucial challenges in seismic data processing is the reduction of the noise in the data or improving the signal-to-noise ratio. In this study, the 1D undecimated discrete wavelet transform (UDWT) has been acquired to attenuate random noise and ground roll. Wavelet domain ground roll analysis (WDGA) is applied to find the ground roll energy in the wavelet domain. The WDGA will be a substitute method for thresholding in seismic data processing. To compare the effectiveness of the WDGA method, we apply the 1D double density discrete wavelet transform (DDDWT) using soft thresholding in the random noise reduction and ground roll attenuation processes. Seismic signals intersect with ground roll in the time and frequency domains. Random noise and ground roll have many undesirable effects on pre-stack seismic data, and result in an inaccurate velocity analysis for NMO correction. In this paper, the UDWT by using the WDGA technique and DDDWT (using the soft thresholding technique) and the regular Fourier based method as f?k transform will be used and compared for seismic denoising.

Coherent and random noise attenuation via multichannel singular spectrum analysis in the randomized domain

ABSTRACTThe attenuation of coherent and random noise still poses technical challenges in seismic data processing, especially in onshore environments. Multichannel Singular Spectrum Analysis (MSSA) is an existing and effective technique for random‐noise reduction. By incorporating a randomizing operator into MSSA, this modification creates a new and powerful filtering method that can attenuate both coherent and random noise simultaneously. The key of the randomizing operator exploits the fact that primary events after NMO are relatively horizontal. The randomizing operator randomly rearranges the order of input data and reorganizes coherent noise into incoherent noise but has a minimal effect on nearly horizontal primary reflections. The randomizing process enables MSSA to suppress both coherent and random noise simultaneously. This new filter, MSSARD (MSSA in the randomized domain) also resembles a combination of eigenimage and Cadzow filters. I start with a synthetic data set to illustrate the basic concept and apply MSSARD filtering on a 3D cross‐spread data set that was severely contaminated with ground roll and scattered noise. MSSARD filtering gives superior results when compared with a conventional 3D f‐k filter. For a random‐noise example, the application of MSSARD filtering on time‐migrated offset‐vector‐tile (OVT) gathers also produces images with higher signal‐to‐noise ratios than a conventional f‐xy deconvolution filter.

Multiple-input adaptive seismic noise canceller for the attenuation of nonstationary coherent noise

Attenuation of coherent noise, typically weather generated noise and more specifically swell noise, is a major concern in seismic. Such noise, usually characterized by its low-frequency content and high amplitudes, is a common problem in seismic acquisition and, in particular, for marine data. We propose a multiple-input adaptive noise canceller as a solution to attenuate nonstationary coherent noise. This filter uses multiple noise sequences to estimate the noise contained in each trace. This noise estimate is then subtracted to obtain an estimate of the trace whose coherent noise has been attenuated. For the implementation of the adaptive filter, we use a variable step-size normalized least mean squares algorithm, as it is known for its simplicity and robustness. In addition, we demonstrate that variable step-size is necessary for this filter to adapt to the changing statistics of the seismic data. This method is tested on real marine seismic data and compared to a time-frequency median filter and a second-order high-pass Butterworth filter. We demonstrate that the multiple-input adaptive noise canceller is a powerful and efficient filter to attenuate swell noise while preserving the seismic reflections. Depending on the noise configuration, it can be used either by itself or in combination with a time-frequency median filter to obtain the best results.

Seismic conditioning and attenuation of high-angle coherent noise in a mixed carbonate and siliciclastic setting, Campos Basin, offshore Brazil: A case study

We demonstrate the use of two industry-standard, noise-attenuating algorithms, [Formula: see text] and [Formula: see text] filtering, to attenuate coherent noise, improve event continuity, and obtain more reliable interpretation within the Pampo field, Campos Basin, offshore Brazil. We analyze 3D seismic data for noise type and associated maximum dip, and then determine optimal attenuating parameters required by the algorithms. In the crossline direction, coherent noise with dips as high as 16 to 18 ms/trace in the data masks real reflected signals. In the inline direction, the noise exhibited an approximately zero dip. The filtering process eliminated noise trains and improved event continuity, as well as obtained better well-to-seismic tie. The output wavenumber associated with noise trains is less than the Nyquist wavenumber; therefore, we conclude that the noise is not related to spatial aliasing, but to other undetermined sources. Though the noise occurs within the entire data volume, it is most severe in the vicinity of the carbonate platform. By eliminating linear noise that masquerades as geologic structures, we obtain better imaging of the Albian succession and the younger Cretaceous turbidite sequences, as well as the top of Aptian salt. The result is a better interpretation.

Attenuation of aliased coherent noise: modelbased attenuation for complex dispersive waves

To achieve an improved understanding of near surface seismic data, Claudio Strobbia, Alexander Zarkhidze, Roger May, John Quigley and Phil Bilsby of WesternGeco present a model-based approach to source-generated coherent noise attenuation.

Coherent noise attenuation in GPR data by linear and parabolic Radon Transform techniques

Ground Penetrating Radar (GPR) is a geophysical method increasingly used in numerous shallow applications. Unfortunately, electronic or acquisition problems can cause the presence in the radargrams of coherent noise interfering with the useful signal. A commonly observed phenomenon, especially for not-shielded antennae, is the surface-scattering effect, due to reflection or diffraction from above-surface objects. These noise events appear with a characteristic hyperbolic moveout in the usual common-offset sections. Other frequent problems are related to the presence of horizontal or dipping features due to system-ringing or other non-geological causes. Several methods have been tried to overcome these problems, most of which involve time domain or Fourier domain filtering. This work presents an attempt to reduce some of these noise modes by an original adaptation of filtering techniques implemented in the Radon domain. The Radon Transform (RT), both in the linear (or t-p) and in the parabolic version (or t-q), has been widely used in seismic processing, especially for multiple removal, but is still quite unfamiliar to GPR practitioners. The results achieved by different RT based methods for coherent noise attenuation in a GPR field example, compared to those of more conventional techniques, are quite encouraging.

Random and coherent noise attenuation by empirical mode decomposition

We have devised a new filtering technique for random and coherent noise attenuation in seismic data by applying empirical mode decomposition (EMD) on constant-frequency slices in the frequency-offset [Formula: see text] domain and removing the first intrinsic mode function. The motivation behind this development is to overcome the potential low performance of [Formula: see text] deconvolution for signal-to-noise enhancement when processing highly complex geologic sections, data acquired using irregular trace spacing, and/or data contaminated with steeply dipping coherent noise. The resulting [Formula: see text] EMD method is equivalent to an autoadaptive [Formula: see text] filter with a frequency-dependent, high-wavenumber cut filtering property. Removing both random and steeply dipping coherent noise in either prestack or stacked/migrated sections is useful and compares well with other noise-reduction methods, such as [Formula: see text] deconvolution, median filtering, and local singular value decomposition. In its simplest implementation, [Formula: see text] EMD is parameter-free and can be applied to entire data sets without user interaction.

Full-azimuth imaging using circular geometry acquisition

Wide-azimuth (WAZ) towed-streamer acquisition has improv-ed the quality and reliability of subsalt imaging in the Gulf of Mexico, where WAZ surveys have typically been ac-quired using 3–4 seismic vessels, each shooting in straight parallel lines. However, acquisition of WAZ data using a single vessel and a circular geometry offers several potential ad-vantages, both for operational efficiency and geophysical analysis of subsurface rock properties. Modeling exercises indicate that circular acquisition geo-metry can deliver a better range of azimuths and offsets than parallel WAZ geometry. A feasibility test in the Gulf of Mexico shows that it is possible to sail along circles while main-taining constant streamer separation and achieving very accurate receiver positioning. Single-sensor recording enables effective attenuation of the additional coherent noise introduced by shooting in a curve. Prestack depth-migrated data from the circular geometry test compare favorably with a parallel geometry WAZ data set...

Local linear coherent noise attenuation based on local polynomial approximation

We propose a new technique for the attenuation of locally coherent noise. We assume that the moveout of the noise is locally linear and approximate its amplitude variations with offset using piecewise (local) polynomial models. Thus, our method consists of three steps: detection of the noise (locally linear coherent noise, LLCN), amplitude estimation by a local polynomial approximation (LPA), and subtraction of the estimated coherent noise from the original data. Applying the proposed method to synthetic data and to a field data set shows that the LPA filter has good ability to model LLCN and is insensitive to the filter parameters. Comparisons of the results obtained by our method with those from the traditional frequency-wavenumber filter and the localized 2D filter in the Fourier projection domain (FPF) show that the new method outperforms both traditional methods in situations with complex coherent noise.

Multi-scale geophysical data analysis using the eigenimage discrete wavelet transform

Discrete wavelet transforms are useful in a number of signal-processing applications. When non-stationary noise spans most wavelet scales, simple rejection of certain scales in the wavelet domain often fails to achieve the desired effect because of strong intra-scale coupling effects. To improve the scale resolution, a joint function of time, scale and eigenvalue that describes the energy density or intensity of a signal simultaneously in the wavelet and eigenimage domains is constructed. A hybrid method, which decomposes eigenimages in the wavelet domain, is developed and tested on field data with a variety of noise types. Several illustrative examples examine the ability of wavelet transforms to resolve features at several scales and the feasibility of combining these transforms with eigenstructure seismic interference cancelling techniques. Successful applications to time-lapse seismic reservoir monitoring are presented. In reservoir monitoring, the scale-dependent properties of the eigenstructure of the 4D data covariance matrix enable us to extract the low-frequency time-lapse signal that is the result of internal diffusive losses caused by fluid flow. The method also gives excellent results in coherent or random noise attenuation, wavefield separation, low-frequency processing, optimal velocity smoothing, data upscaling, and integration of well logs and seismic data. Particular benefits are due to trace matching for multiple subtraction and 4D cross-equalization that extracts subtle variations due to changes in reservoir properties from a signal whose spectral density is not stationary.

Beamforming with the generalized sidelobe canceller for coherent noise attenuation in 3D seismic data

This paper presents a new approach for attenuating coherent noise in 3D seismic data. An adaptive beamforming with generalized sidelobe canceller (GSC) design methodology is utilized here as a general form of linearly constrained adaptive beamforming structure. It consists of a fixed beamformer, and a signal-blocking matrix in front of an unconstrained adaptive beamformer. Considerationf of the complexity of the geometry for 3D seismic survey, the 3D beamforming with GSC technique is developed with two key points: (1) sorting along azimuth sections to simplify the relationship between traveltime and offset from 3D to 2D, and (2) dynamic binning scheme to avoid the possible poor folding in some azimuth sections. Both simulation result and real data example show that the newly developed 3D beamforming with GSC yields more credible results at a relative low cost, sufficient stability and good resolution.

Improvement in GPR coherent noise attenuation using τ‐p and wavelet transforms

We present a new application of modern filtering techniques to ground‐penetrating radar (GPR) data processing for coherent noise attenuation. We compare the performance of the discrete wavelet transform (DWT) and the linear Radon transform (τ‐p) to classical time‐space and Fourier domain methods using a synthetic model and real data. The synthetic example simulates problems such as system ringing and surface scattering, which are common in real cases. The field examples illustrate the removal of nearly horizontal but variable‐amplitude noise features. In such situations, classical space‐domain techniques require several trials before finding an appropriate averaging window size. Our comparative analysis indicates that the DWT method is better suited for local filtering than are 2D frequency‐domain (f‐f) techniques, although the latter are computationally efficient. Radon‐based methods are slightly superior than the techniques previously used for local directional filtering, but they are slow and quite sensitive to the p‐sampling rate, p‐range, and sizes of the muting zone. Our results confirm that Radon and wavelet methods are effective in removing noise from GPR images with minimal distortions of the signal.

Interpolation and multiple attenuation with migration operators

A hyperbolic Radon transform (RT) can be applied with success to attenuate or interpolate hyperbolic events in seismic data. However, this method fails when the hyperbolic events have apexes located at nonzero offset positions. A different RT operator is required for these cases, an operator that scans for hyperbolas with apexes centered at any offset. This procedure defines an extension of the standard hyperbolic RT with hyperbolic basis functions located at every point of the data gather. The mathematical description of such an operator is basically similar to a kinematic poststack time‐migration equation, with the horizontal coordinate being not midpoint but offset. In this paper, this transformation is implemented by using a least‐squares conjugate gradient algorithm with a sparseness constraint. Two different operators are considered, one in the time domain and the other in the frequency‐wavenumber domain (Stolt operator). The sparseness constraint in the time‐offset domain is essential for resampling and for interpolation. The frequency‐wavenumber domain operator is very efficient, not much more expensive in computation time than a sparse parabolic RT, and much faster than a standard hyperbolic RT. Examples of resampling, interpolation, and coherent noise attenuation using the frequency‐wavenumber domain operator are presented. Near and far offset gaps are interpolated in synthetic and real shot gathers, with simultaneous resampling beyond aliasing. Waveforms are well preserved in general except when there is little coherence in the data outside the gaps or events with very different velocities are located at the same time. Multiples of diffractions are predicted and attenuated by subtraction from the data.

Coherent noise attenuation in the radial trace domain

Coherent noise is a persistent problem in seismic imaging, and a number of techniques have been developed to attenuate it. The radial trace (RT) transform, a simple seismic data mapping algorithm, can be used as the basis for a particularly flexible and effective method for attenuating coherent noise on both prestack and poststack seismic data. Described here are the principles and some practical application details for attenuating coherent noise in the RT domain. A comparison between frequency–wavenumber (f–k) and RT domain filtering on a synthetic model is presented, and some of the differences and advantages of RT methods are identified. Next, RT coherent noise attenuation is demonstrated using a set of good‐quality field data; it is then applied to a very noisy data set. The results obtained with this last set prove to be as good as, or better than, those produced using f–k filtering.

Perturbations in 4D marine seismic

A new concept for acquiring calibrated towed streamer seismic data is introduced through a new acquisition and processing system. The specification of the new system has been defined by rigorous analysis of the factors that limit the sensitivity of seismic data in 4D studies and imaging. New sensor and streamer technology, new source technology and advances in positioning techniques and data processing have addressed these limitations.Sensitivity analysis revealed that the most significant perturbations to the seismic signal are swell noise and sensor sensitivity variations. Conventional analog groups of hydrophones are designed to suppress swell noise however a new technique for data-adaptive coherent noise attenuation delivers even greater noise suppression for densely spatially sampled single-sensor data. Although modern source controllers provide accurate airgun firing control the signature of an airgun array may vary from shot to shot. This can be due to factors such as changes in the array geometry, air pressure variations, depth variations and wave action. A method for estimating the far-field signature of a source array is the Notional Source Method (proprietary to Schlumberger) which has been steadily refined since its first disclosure. A recent development compensates for variation in source array geometry by monitoring the position and azimuth of each subarray using GPS receivers mounted on the floats.New calibrated positioning and streamer control systems are part of the new acquisition system. Active vertical and lateral streamer control is achieved using steerable birds and positioning uncertainty is reduced through an in-built fully braced acoustic ranging system. Calibrated marine seismic data are achieved through quantifying the source output, the sensor responses and positioning uncertainty. The consequential improvements in seismic fidelity result in better imaging and more reliable 4D analysis.