Due to the absorption of the formation, the frequency band of seismic wave narrows in the process of propagation and the resolution decreases. Spectral modelling deconvolution improves the resolution of seismic data by fitting the wavelet amplitude spectrum and applying zero-phase deconvolution on the seismic signal. Considering the different frequency components of seismic data from different depths, compensating the spectrum differently according to the depth of seismic data is meaningful, in which the time-frequency spectrum is needed. Recently proposed time-frequency analysis method named Synchrosqueezed Wavelet Transform (SSWT) is superior to the traditional time-frequency analysis method in time-frequency resolution, which has been applied to seismic data processing and interpretation. In this paper, we propose a method named Time-Varying Spectral Modeling Deconvolution Based on SSWT and apply it to seismic data to improve the resolution. The effectiveness of the proposed method in improving the resolution is verified by testing field data and comparing it with the results based on Generalized S-transform.
Presentation Date: Thursday, September 28, 2017
Start Time: 9:45 AM
Presentation Type: ORAL
Ground roll is one of the regular interference waves in seismic data. The removal is difficult when ground roll is scattered, and data quality is often impaired. Traditional suppression methods such as the f-k filtering often lose effectiveness because their energy is distributed over a wide portion of the f-k spectrum. However, the EMD algorithm can separate the ground roll directly through its orthogonality and dyadic spectral characteristics, but it is difficult to locate the ground roll component. In this paper, we modified the EMD algorithm with average value constraint and proposed a new method, which takes the weak changes of strong low-frequency signals into account when evaluate the envelope. Numerical results have been performed in field data to illustrate the efficiency of our method.
Presentation Date: Tuesday, October 18, 2016
Start Time: 9:15:00 AM
Presentation Type: ORAL
Due to the existence of anisotropy and viscosity, seismic wave in the propagation process shows the difference in all directions and attenuation characteristics, which seriously affect the effect of migration. In this paper, based on a single SLS isotropic media theory, we extend it for anisotropic medium to get the viscoacoustic wave equation in TTI medium, and applied it achieve reverse time migration of anisotropic medium. HESS model shows directly that reverse-time migration imaging based on anisotropic viscoacoustic wave equation can produce clearly imaging and improve amplitude beneath highattenuation zones, and balanced amplitude. So we get accurate and reliable amplitude preservation imaging section.
Alkhalifah (1998) first proposed the pseudo-acoustic wave equation in transversely isotropic media, which was derived from P-SV wave dispersion relation that S-wave velocity along the symmetry axis is set to be zero. It includes fourthorder partial derivatives of the wavefield in time and space equation, which leads to complexities in numerical implementation; Zhou et al. (2006) and Du et al. (2008) simplified the dispersion relation by introducing an auxiliary wavefield, and decomposes the fourth-order differential equation into two coupled second-order partial differential equations, which are more convenient to solve numerically. By reason of the instability for the residual shear, Fletcher et al. (2009) proposed adding nonzero Swave velocity terms to overcome the problem. Duveneck et al. (2011) proposed a TTI acoustic formulation according to the anisotropic elastic equations. Zhang et al. (2012) pursued a stable TTI acoustic wave equations by introduce self-adjoint differential operators and proved the new equations are stable in the sense that energy of the wavefields is conserved during the propagation.
Besides, the effects of viscosity can cause seismic energy attenuation and wavelet distortion, thus greatly reduce resolution in the image. Zhang et al. (2010) derived isotropic viscoacoustic wave equation with pseudodiffererntial operator in time domain, and applied it in reverse time migration. Suh et al. (2012) extended Zhang’s method to anisotropic media medium, compensating for attenuation effects of anisotropic media. Based on a single SLS, Bai et al. (2012 and 2013) derived a viscoacoustic wave equation for wave propagation and its adjoint. The equation doesn’t involve any memory variable, so it can solve the instability owing to without memory variable of the recorded data.
The underground medium is far from being isotropic and elastic. Such simplifications in modeling the seismic response of real geological structures may lead to misinterpretations, or even worse, to overlooking useful information. The existence of anisotropy and viscosity affect the effect of migration seriously. So, in this paper, based on the dissipation mechanism of standard linear solid model, we derive a pure viscoacoustic wave equation of TTI media in the time domain, which describes the attenuation characteristics of seismic wave by a Pseudodifferential operators in the equations. The results of numerical simulation show that the equation propagation can not only describe the propagation of pure scalar wave in anisotropic media accurately, but also reflect the effect of absorption and attenuation.
At present, there are two kinds of methods to implement scalar wave RTM in anisotropic medium, those are pseudoacoustic wave RTM and pure acoustic wave RTM. At first, Alkhalifah (1998) first proposed the pseudo-acoustic wave equation in transversely isotropic media, which includes fourth-order partial derivatives of the wavefield in time and space equation. Zhou et al. (2006) and Du et al. (2008) simplified the fourth-order differential equation into two coupled second-order differential equations. However all of this does not eliminate the effect of shear wave completely. By reason of the instability for the residual shear, Fletcher et al. (2009) proposed adding nonzero S-wave velocity terms to overcome the problem. Duveneck et al. (2011) proposed a TTI acoustic formulation according to the anisotropic elastic equations. Zhang et al. (2012) pursued a stable TTI acoustic wave equations by introduce selfadjoint differential operators and proved the new equations are stable in the sense that energy of the wavefields is conserved during the propagation.
On the other hand, Liu et al. (2009) derived the pure Pwave and SV wave equation by solving the coupled equations. Reynam et al. (2011), Ge Zhan et al. (2011) give the pure P-wave and pure SV wave equation by solving the decoupling of the dispersion relation P and SV waves. Owning to the huge amount of calculation in the spatial domain, they implemented it in the frequency-wavenumber domain. When the anisotropy parameter variation is not intense, Chu (2013) proposed numerical interpolation scheme, which can solve instability problems caused by anisotropy parameter changing in spatial domain. Xu et al. (2014) proposed a new algorithm of pure quasi-P wave equation, which decomposes the pseudo-differential operator into two solvable operator: one Laplacian operator and one scale operator, and applied it for reverse time migration in anisotropic media.
Zhao, Ying (China University of Petroleum, Huadong) | Yue, Youxi (China University of Petroleum, Huadong) | Huang, Jianliang (CNOOC Research Institute) | Wang, Jiao (China University of Petroleum, Huadong) | Liu, Bingqing (China University of Petroleum, Huadong) | Liu, Chenxi (China University of Petroleum, Huadong)
It is important to detect the discontinuity of seismic data for the identification and depiction of underground abnormal geologic body. In this paper, we introduce three parameters wavelet (TP wavelet) with higher time-frequency resolution and better energy focusing nature and propose a multi-scale edge detection method of structure-oriented Sobel gradient attribute based on TP wavelet transform. This method uses the dip and azimuth information as constraints and realizes multi-scale discontinuity detection of any profile or slice in 3D seismic data. Application of this new method to real seismic data shows it can effectively eliminate the interference of background formation information and clearly detect abnormal geological boundaries. Moreover, it can reflect richer detail information and has an advantage in effectively identify and finely characterize different scale of seismogeological edges. Thus our method can be very useful in the identification of horizontal difference of layers and the edge of geologic bodies.
In the seismic data, the characteristics of discontinuity such as faults and channel sand boundaries are reflected as the edge features in the image. The application of edge detection technique to these edges can effectively image geological phenomena contained in data and make the identification of geological features easier. It is helpful for interpreter to make more effective and accurate geological interpretation. Currently, there are lots of edge detection methods, such as the classic operators, wavelet multi-scale edge detection algorithm and so on. These methods are essentially based on the 2D thought and detect 2D seismic images. It will not get good results if they are used directly to identify unusual geological boundaries in 3D seismic data. For discontinuity detection of 3D seismic data, Aqrawi et al. (2011) used firstly cross-correlation method to get dip formation, then applied dip-constraint Sobel operator to the detection of fault and salt dome boundaries and achieved very good results, but he didn’t give a specific operator implementation scheme. Song et al. (2013) used Aqrawi’ idea for reference and based on the estimate of dip and azimuth formation by gradient structure tensor, then proposed specific implementation scheme of structureoriented Sobel operator. Since geological edge has multiscale features and three-parameter wavelet has better timefrequency focusing property than Morlet wavelet and the best matching the seismic wavelet (BMSW), we put forward a new multi-scale edge detection method of structure-oriented Sobel gradient attribute based on threeparameter wavelet’ multi-scale analysis. Application this method to real seismic data shows good results.