超深层复杂构造地震波高精度成像技术数据

逆时偏移（RTM）技术采用双程波波动方程进行波场延拓，避免对波动方程的近似，具有适应复杂记录波场、复杂地下构造以及可以成像回转波、棱柱波、多次波等优点。但是，在对三维数据进行逆时偏移成像时，就存在计算量大，计算效率低的问题，将三维自适应变网格正演模拟算子移植入逆时偏移技术，就可以提高成像效率。采用面向目标的最小二乘逆时偏移，通过设置不同的加权函数对数据残差中的反射波进行加权，可以实现面向不同地质目标的高精度成像。在对绕射波成像时，该方法可以通过迭代更新算法改善绕射波场的成像质量，提供地下非均质地质体的精确图像。在对全波场成像时，该方法在反射波场加权函数逐渐减小的过程中，首先更新反射波场对应的大尺度均质构造，改善其成像质量的同时逐渐减小反射波场在目标函数中的比重，突出绕射波场，以改善小尺度的非均质构造的成像质量，最终可以得到地下所有构造的高精度图像。奇异谱分析最初的引入是用于地震记录的去噪及重构线性信号，基于降秩的手段以分离噪音及强线性同相轴。但当我们把它引入到绕射波分离，由于其本身的缺陷使得线性能量发生延拓，在与原始信号匹配做差值而获得的绕射信号，其反射层边界处能量因失配而发生相位反转。因此奇异谱分析无法用于局部变化的反射层，即断层等绕射信息丰富的地区，与适用于横向变化区域的逆时偏移的目的相矛盾。我们提出了短时奇异谱分析方法，即在短时傅立叶变换中实现奇异值分解，以达到良好的绕射波分离及去噪的效果，其STSSA作用在平面波最小二乘逆时偏移中的合成平面波道集上；根据相速度频散关系，推导了各向异性TTI介质中的波场传播算子，并发展了相应的各向异性TTI介质逆时偏移成像方法。该方法可考虑倾斜沉积层及高陡构造引起的波形畸变，使得反传波同相轴得以准确归位，绕射波准确收敛，提高了偏移成像分辨率。

Reverse Time Migration (RTM) adopts two-way wave equations for wavefield continuation, avoiding approximations to the wave equation. It boasts advantages including adaptability to complex recorded wavefields and complex underground structures, as well as the capability to image turning waves, prism waves and multiple waves. However, RTM imaging on 3D data suffers from large computational cost and low efficiency. Incorporating a 3D adaptive variable-grid forward modeling operator into RTM can effectively improve imaging efficiency.

Goal-oriented least-squares reverse time migration enables high-precision imaging targeting specific geological objectives by applying different weighting functions to weight reflected waves in data residuals. For diffracted wave imaging, this method can enhance the imaging quality of diffracted wavefields via iterative update algorithms, delivering accurate images of underground heterogeneous geological bodies. For full-wavefield imaging, as the weighting function for reflected waves gradually decreases, the method first updates large-scale homogeneous structures corresponding to reflected wave fields. While improving their imaging quality, it gradually reduces the proportion of reflected wave fields in the objective function, highlighting diffracted wave fields to optimize the imaging quality of small-scale heterogeneous structures, ultimately yielding high-precision images of all underground structures.

Singular Spectrum Analysis (SSA) was originally introduced for seismic record denoising and linear signal reconstruction, separating noise and strong linear events through rank-reduction techniques. However, applying SSA to diffracted wave separation introduces inherent defects: linear energy undergoes continuation, and the diffracted signal obtained by subtracting the matched original signal exhibits phase inversion at reflection layer boundaries due to energy mismatch. Consequently, SSA is inapplicable to laterally varying reflection layers, such as fault zones rich in diffracted information, which conflicts with the application scenario of RTM suitable for laterally varying regions.

We propose the Short-Time Singular Spectrum Analysis (STSSA) method, which implements singular value decomposition within the short-time Fourier transform to achieve excellent diffracted wave separation and denoising performance. STSSA is applied to synthetic plane-wave gathers from plane-wave least-squares reverse time migration. Based on the phase velocity dispersion relation, we derive the wavefield propagation operator for anisotropic TTI media and develop the corresponding RTM imaging method for such media. This method accounts for waveform distortions caused by tilted sedimentary layers and high-steep structures, enabling accurate positioning of back-propagated wave events and precise convergence of diffracted waves, thereby improving the resolution of migration imaging.

搜集汇总

背景与挑战

背景概述

该数据集提供了超深层复杂构造地震波的高精度成像技术数据，采用多种先进技术方法如逆时偏移和最小二乘逆时偏移，以提高成像效率和精度，适用于勘探地球物理学领域的研究。数据量为26.43GB，共72个文件。