Layer | Fill | Outline |
---|

#### Map layers

Theme | Visible | Selectable | Appearance | Zoom Range (now: 0) |
---|

Fill | Stroke |
---|---|

### Collaborating Authors

ABSTRACT The attenuation of source-generated coherent noise energy can be a challenging problem for land data where surface waves often exhibit complex behavior with multiple propagation modes, high lateral variability and relatively short wavelengths. The traditional acquisition and processing strategy for mitigation of coherent noise has combined analog spatial filtering through source and receiver arrays in the field, with multi-channel digital filtering in data processing. The field arrays act as complementary spatial anti-alias filters for data processing algorithms which have difficulty in dealing with aliased events. Limitations of the available processing procedures place constraints on the acquisition design which can potentially both limit flexibility and increase the cost of the acquisition. A new model-based approach to source-generated coherent noise attenuation is presented, where the local properties of the multi-mode surface waves are estimated from the seismic data and used to generate a detailed model of surface-wave noise, spatially-variable over the survey area. The method has significant advantages with respect to the handling of aliased coherent noise energy, and robustness to spatial irregularities. The availability of effective processing tools for aliased noise attenuation can have a significant impact on required survey geometry, and on the cost of land exploration.

- Asia > Middle East (0.19)
- Africa > Middle East (0.16)

- Geophysics > Seismic Surveying > Seismic Processing (0.97)
- Geophysics > Seismic Surveying > Seismic Modeling > Velocity Modeling (0.48)

Although many acquisition and processing techniques are successful in suppressing this type of noise for conventional 2-D where is defined by the product of time data, such methods are often ineffective for the 3-D case. This delay and advance operators, a), and a. weighting function, presentation discusses an approach to coherent-noise suppression Such a minimization scheme is independent of s(w,

Summary Some conventional noise attenuation methods in seismic data processing often need assume that coherent events are piecewise-stationary, piecewise-linear and regularly sampled along spatial direction. In this paper, without any spatial assumptions about coherent signal, a noise attenuation method using Bayesian inversion is presented. Its essential idea is to directly invert the โcleanโ data, regarded as model parameter, from observed seismic data by maximizing a posterior distribution, which is made up of prior distribution and likelihood function. Whether this method can reduce noise is dependent on the choice of prior information. Based on a statistical knowledge that coherent data oscillates slightly and random noise strongly, the minimization of L1 norm of model parametersโ difference quotient, also called as total variation, is used as prior information. What advantage of this method is that it can enhance nonstationary and nonlinear seismic events. Moreover, the de-noised effect neither strongly relies on the size and layout of time windows nor depends on whether traces are sampled regularly. Especially, this method has good ability for preserving edges of discontinuous events, which often correspond to important geologic features, and deblurring amplitudeโs variation along spatial direction, which is probably AVO response. A model data and a real data are used to test its validity. Introduction The structural feature and physical property of subsurface reflectors can be reflected in seismic section. Thatโs just what geophysicists make use of to explore oil and gas. Unfortunately, besides effective response from reflectors, some random disturbance is also recorded in seismic sections, which compromises our ability for describing the underground medium. Generally speaking, conventional random noise attenuation methods can be classified into two categories: (1) prediction filtering method in the frequency-space (f-x) domain, mainly including f-x deconvolution (Canales, 1984), and (2) eigen-decomposition-and-reconstruction method, mainly including (local) singular value decomposition (SVD) filtering (Ulrych et al., 1988; Bekara and van der Baan, 2007) and (partial) Karhunen-Loeve transform filtering (Jones and Levy, 1987; Al-Yahya, 1991). Both this two methods are based on the fact that coherent events and noise can be separated in certain domain. However, this separability is strictly limited by some conditions or assumptions. For prediction filtering, we need assume that coherent events are linear and stationary, and the trace interval is equal. For eigen-decomposition-andreconstruction, we need assume that coherent events are horizontal (dipping events should be flattened by dip steering), the amplitude of one trace for any event is proportional to that of any other traces, the phase is invariant among traces, and there are no conflicting events in the section. However, real data does not satisfy all assumptions of every method, thus the de-noised result is often not as ideal as we expect. Even if some techniques, such as sliding time windows and dip steering, are applied to try to make seismic data meet those geometry assumptions, they will make noise-reduction process complicated and bring other new problems. In this paper, noise attenuation using Bayesian inversion is proposed, which does not need any spatial assumptions about coherent events.

ABSTRACT No preview is available for this paper.

- Geophysics > Seismic Surveying > Seismic Processing > Seismic Migration (0.96)
- Geophysics > Seismic Surveying > Seismic Modeling > Velocity Modeling (0.70)

Summary We examined the robustness of the Laplace-domain waveform inversion for coherent noise. Numerical examples of two synthetic data sets contaminated by linear and hyperbolic coherent noise demonstrate that the Laplace-domain waveform inversion is robust for coherent noises. Through the analysis of the Laplace-transformed wavefields, we can confirm that coherent noise have only insignificant effect on Laplace-domain wavefields. Therefore, the Laplace-domain waveform inversion can be used to obtain reasonable long-wavelength velocity models even from data containing coherent noise, the marine seismic data believed to be contaminated by elastic waves such as P-S converted waves. Introduction Full waveform inversion is a method of estimating the subsurface velocity structure that is more advanced than stacking velocity analysis, migration velocity analysis, or traveltime tomography. Generally, the method of estimating a velocity model via full waveform inversion suffers from random or coherent noise. Coherent noise severely degrades the quality of the subsurface information that can ultimately be delineated from the data (Linville and Meek, 1995). For inverse problems, noise makes the problems non-unique or ill-posed (LaBrecque et al., 1996; Parker, 1984). Although many algorithms are influenced by noise, full waveform inversion is also believed to be sensitive to all kinds of noise. Noise can be reduced to some degree by adequate preprocessing. However, heavy pre-processing results in the distortion of the original amplitude and phase, deteriorating the inversion results. Although many researchers have introduced robust objective functions, such as the Cauchy criterion, the hyperbolic secant, the l1-norm (Crase et al., 1990; Djikpesse and Tarantola, 1999), and the Huber function (Bube and Langan, 1997), they are only valid for outliers or missing traces, not for coherent noise. Recently, Shin and Cha (2008) proposed a robust waveform inversion algorithm that can recover long-wavelength velocity structure. They exploited the advantages of Laplace-domain waveform inversion to produce long-wavelength velocity models even from data without low frequency components. However, the robustness of Laplace-domain inversion to real marine seismic data believed to have mode converted P-S or S-P waves has not been fully explained. In this paper, we address why Laplace-domain waveform inversion is robust for coherent noise by showing examples of the inversion of data that are contaminated by coherent noise. For the analysis of sensitivity to noise, we perform Laplace-domain inversion using two synthetic data sets, one of which contains linear coherent noise while the other contains hyperbolic coherent noise. Review of Laplace-domain waveform inversion Before discussing the robustness of the Laplace-domain waveform inversion to real data, we briefly review the Laplace-domain waveform inversion algorithm proposed by Shin and Cha (2008). We start with the definition of the Laplace-domain wavefield, and then end with the inversion algorithm. The wavefield in the Laplace domain can be expressed as where s is the Laplace damping constant, U( x , s) is the Laplace-domain wavefield, and u( x ,t) is the time-domain wavefield. Since the time-domain wavefield is a causal signal, the interval of integration in Equation (1) can be replaced by the interval from negative infinity to positive infinity. Then, we can rewrite Equation (1) as