论文部分内容阅读
本文研究分析了双程波波动方程偏移成像中广泛存在的三种主要噪声,特别是针对过去研究中没能很好解决的存在于高速盐丘悬垂边界附近的射线状噪声,提出了基于优化成像条件的有效去噪方法。射线状噪声主要来自于震源一侧波场的下行透射波分量和接收阵列一侧波场的上行散射波互相关成像。这一部分能量具有较强的互相关性,但并不携带真实的反射面信息。它广泛存在叠前偏移成像中,与信号在强度上同量级。多数情况下偏移成像中的相关噪声由方向性传播的波场能量产生。利用波场梯度得到的波场传播角度,可以分离出噪声对应的波场能量,并在成像条件中减去。采用这一方法可以有效地去除多种噪声,包括直达波噪声、散射波噪声和射线状噪声。该去噪方法不依赖波场外推算子,在需要时可以方便地运用到几乎所有的波动方程偏移中去。并且该去噪方法针对噪声的物理根源,对信号的损害很小。对去噪后的偏移成像结果额外地进行波数域滤波处理,可以进一步提高叠加图像的质量。这一去噪方法在超广角单程波偏移成像中取得良好效果,我们同时期待其在其他双程波波动方程偏移特别是逆时偏移(RTM)中的成功运用。
In this paper, we analyze and analyze the three major types of noise that exist widely in the migration imaging of the two-way wave equation. Especially for the ray-like noise that exists in the vicinity of the hanging boundary of the high-speed salt dome that was not well solved in the past, An effective denoising method for imaging conditions. Radiated noise mainly comes from the cross-correlation imaging of down-going transmitted wave components in the wave field of the source side and the up-scattered wave of the wave field in the receiving array side. This part of the energy has a strong cross-correlation, but does not carry the true reflective surface information. It exists widely in prestack migration imaging, and is of the same order of magnitude as the signal. In most cases, the correlated noise in the offset imaging is generated by the wavefield energy that is directionally propagated. Using the wave field propagation angle obtained from the wave field gradient, the wave field energy corresponding to the noise can be separated and subtracted from the imaging conditions. This method can effectively remove a variety of noise, including direct wave noise, scattered wave noise and radio noise. The denoising method does not depend on the wave field extrapolation operator and can be easily applied to almost any wave equation migration when needed. And the denoising method is aimed at the physical origin of the noise and has little damage to the signal. The wave number domain filtering process is additionally performed on the denoised offset imaging result to further improve the quality of the superimposed image. This de-noising method works well in ultra-wide-angle one-way migration imaging and we also look forward to its successful use in other two-way wave migration equations, especially reverse time migration (RTM).