The accurate beam migration method without slant stack under dual-complexity conditions and its application (I) : Acoustic equation
Huang Jian-Ping,Yuan Mao-Lin,Li Zhen-Chun,Yue Yu-Bo
DOI: https://doi.org/10.6038/cjg20150124
2015-01-01
Abstract:Recent trends in seismic exploration have transferred to lands where both topography and subsurface targets are complex, leading to challenges for both acquisition and processing methods. Compared with one-way or two-way wave-equation migration, ray-based migration is an efficient and flexible method, especially for complex topography. In general, Gaussian beam migration (GBM) is more accurate than single-arrival Kirchhoff migration, although there are some issues resulting in dissatisfactory GBM images. For example, static correction will lead to the distortion of wavefields when near-surface elevation and velocity vary drastically. Worse still, unbalanced amplitude will emerge in migrated images when receivers are not placed within some neighborhood of the beam center, that is, GBM is slightly inflexible for the irregular acquisition system. For the irregular surface, both near-surface velocity and takeoff angle of an individual beam at the beam center are different from those at each receiver, so GBM usually produces non-ignorable migration artifacts. In order to improve the flexibility and accuracy of GBM, we propose an accurate beam migration method without slant stack based on acoustic wave equation. An accurate acoustic beam migration algorithm based on surface dip angle information for complex topography is presented. Unlike the conventional GBM migration method, the proposed method shoots beams directly from receivers without elevation static correction, phase correction, as well as approximate velocity and takeoff angle of an individual beam. Based on acoustic wave equation, we first derive the acoustic backward-continued wavefields from receivers by using surface dip angle information and then derive the acoustic beam migration formula by using the deconvolution imaging condition. Finally, we use a real data and two synthetic datasets separately from the 2D Canadian Foothills model and the Zhongyuan oilfield fault model to verify the proposed algorithm. In the example of 2D Canadian Foothills model, our method suppresses the near-surface noise. Meanwhile, it images near-surface, extremely steep, and overturned structures more clearly, compared to other methods such as wave-equation migration, the Gray' s method and Yue' s method. Furthermore, our method efficiently eliminates imaging energy error caused by the large distance between the beam centre and receivers in both Gray' s method and Yue' s method, and thus improves the imaging amplitude. In the example of Zhongyuan oilfield fault model, the image generated by our method shows fewer migration artifacts and better continuity for the reflectors than those generated by the Yue's method. Our method is capable of producing a higher resolution of images and better S/N of profiles. In the example of real data, compared to the Yue's method, our method produces more continuous reflectors on both flanks of the buried hill and sharper faults on the left side of the area. Moreover, our method generates more interpretable image profiles with more balanced amplitude, especially in the deep formation, while the Yue's method can hardly accomplish this because of the amplitude error caused by the large distance between the beam center and receivers. One feature of our method is that beams for backward-continued wavefields are emitted directly from the receivers, without implementing the following procedures: (1) elevation statics; (2) phase correction; ( 3) approximation of both velocity and takeoff angle of an individual beam between receivers and the beam centre. While in the traditional beam migration methods, those procedures will lower the imaging quality, especially when the near-surface velocity and elevation vary drastically. Compared to wave equation migration (WEM) and other ray-based methods, our method is more effective in suppressing the near-surface noise and improving imaging quality on the near-surface, extremely steep, and overturned structures. Moreover, our method effectively reduces the imaging energy error that is caused by the large distance between the beam centre and receivers, which usually occurs in conventional beam migration methods, and has great potential in improving amplitude and S/N. However, our method is relatively costly because it requires shooting beams from each receiver. We will develop the algorithm for elastic wave equation and then apply it to multicomponent data processing.