Surface-consistent deconvolution is a useful tool to improve the resolution of seismic data by com-pressing the seismic wavelets and enhancing the wavelets’waveform consistency. However，the conventional surface -consistent deconvolution mainly uses the least -squares method or the Gauss -Seidel method for spectral decomposition. Their decomposition results are easily affected by noise，which results in an increase of noise energy in the seismic data after deconvolution. Therefore，this paper proposes the three -dimensional（3D）sur- face-consistent deconvolution method based on the over-relaxation Jacobian iteration algorithm to improve the anti-noise ability of the surface-consistent deconvolution. First，the logarithmic power spectrum is effectively de- composed into five components，including source，receiver，common midpoint（CMP），offset and global term， by using the weighted least-squares objective function under strong noise conditions. Then，the five-component convolution model with the global term is used instead of the conventional four-component model，and the changing near -surface conditions are transformed into surface conditions similar to the global term，which thus effectively eliminates the waveform changes caused by inconsistent near -surface conditions. Test results of two sets of synthetic data and one set of 3D field data show that the over-relaxation Jacobian iteration algorithm has relatively strong anti-noise ability. The algorithm can effectively compress the seismic wavelets，enhance the wavelets’waveform consistency，effectively compress non-surface-consistent noise，and has achieved rela- tively good deconvolution performance，which shows great significance for enhancing the vertical resolution of seismic data in complex geological structures.

The seismic data of the Lower Cambrian Qiongzhusi Formation in the Sichuan Basin has weak re- flected energy without distinct characteristics of interlayer wave groups and clear description of faults，which makes it difficult to meet the requirements for fine prediction of high-quality shale reservoirs and the accurate de- sign of horizontal well trajectories. In response，taking the 3D seismic project for shale gas in Well Z201 as an example，this paper proposes a high-precision seismic acquisition technology of deep shale gas in Qiongzhusi Formation of Sichuan Basin. First，the observation system parameter optimization technology for pre -stack in- version of reservoirs is applied to design the observation system. Then，the intelligent layout of physical points in the obstacle area based on the contribution degree is conducted to improve the uniformity of coverage times in the target zone. Finally，the surface velocity and lithology constrained modeling technology is used to character- ize the near-surface structure and spatial distribution of lithology within the survey area. The application results of the proposed method in seismic acquisition of shale gas in Qiongzhusi Formation demonstrate that high -reso- lution，broadband seismic data can be obtained using the technology，which provides a data basis for the subse- quent high-resolution processing and fine reservoir description.

In ocean bottom node（OBN）seismic exploration，significant discrepancies often exist between the actual node positions and the initial surveyed locations due to factors such as ocean currents，tides，topographic variations，and fishing vessel dragging，necessitating secondary positioning. To address this，a high-precision secondary positioning method based on four-quadrant stacking is proposed. Specifically，within a relative coordi- nate system，shot points are divided into four quadrants according to their azimuth. Linear moveout correction and common receiver point stacking are performed separately for shot points in each quadrant. Subsequently， the displacement of the node from the surveyed position to the actual position is decomposed into two mutually perpendicular components. These two components are determined by analyzing the first arrival time differences from the common receiver points in each quadrant，which thereby enables precise estimation of the node’s ac- tual coordinates. Application to field data demonstrates that this method achieves high -precision secondary posi- tioning of nodes. Moreover，the four -quadrant stacking strategy significantly improves the reliability of first ar- rival picking. Compared with traditional methods，this approach offers higher positioning accuracy and effi- ciency，showing excellent engineering applicability and promising potential for broader adoption.

Diffraction imaging is an important method to improve the imaging accuracy of underground small-scale geologic bodies. However，conventional seismic surveys are mainly based on reflection imaging，with weak-energy diffraction suppressed，as a result of which diffraction need to be separated and imaged separately. At present，the localized damped rank-reduction with adaptively chosen ranks（LDRRA）method widely used improves the separation accuracy of diffraction by damping the adaptively chosen singular value matrix with a damping operator，but its damping factor is mainly given manually，and all local window data use the same given damping factor. Different local windows contain different seismic data，and using the same damping factor will reduce the separation accuracy of diffraction. Therefore，a localized adaptive damped rank-reduction （LADRR）method for diffraction separation is proposed. First，based on the LDRRA framework，the Hankel matrix undergoes singular value decomposition（SVD）to truncate singular values. Second，a squared ratio of singular value is introduced to adaptively compute a damping factor for each localized data window，through which the optimal damping factor is selected to apply damping effects to the truncated singular values，thereby preserving the reflection components. Finally，the damped localized window data is subjected to inverse Hankel- ization and inverse Fourier transform，and then subtracted from the original wavefield to yield the separated dif- fraction. Theoretical simulation and field data test results demonstrate that the proposed method can separate diffraction with high accuracy，and the imaging results of the separated diffraction can get more accurate loca- tion of the underground small-scale geologic body.

With the development of deep learning（DL）techniques，multimodal learning（MML）has been widely applied to solve geophysical inversion problems. Currently，the electromagnetic data with different fea- ture attributes are usually grouped as a whole input to the network when DL is used to invert controlled -source electromagnetics（CSEM）data，which can be regarded as a single-modality learning strategy. MML involves extracting unique features from various modalities of data and leveraging associated information to establish an input-output mapping relation. In this study，electromagnetic data of different frequencies are used as different modalities，and MML is combined with UNet to realize 2D inversion of frequency -domain marine CSEM data. The inversion results on a test set of synthetic data show that the proposed method can accurately reconstruct the resistivity structure of seafloor media，determine the location of seafloor high-resistance bodies，and map their distribution. Compared to traditional inversion methods，this approach is notably efficient and stable. In ad- dition，its inversion performance on noisy data indicates that it has the potential to be applied to field data.

Channel identification is crucial for predicting fluvial facies reservoirs. However，when the P-wave impedance contrast between channel sandstones and surrounding rocks is minimal，it is difficult to use only poststack P -wave seismic data for channel identification. S -wave data can effectively enhance the reliability of predicting the spatial distribution of channels. However，the combined identification process of P-wave and S-wave involves challenges such as difficult parameter selection，high subjectivity，and extended working cycles， leading to inefficiencies and reduced reliability. This paper proposes an automatic channel identification methodbased on the joint P-wave and S-wave seismic data. First，to address the issue of insufficient sample data，it puts forward a method for automatically generating synthetic forward modeling samples of 3D channel geological models based on actual data interpretation and channel interpretation results，effectively expanding the sample data set. Subsequently，a new 3D automatic channel identification network structure is then designed， which effectively integrates P-wave and S-wave seismic data，enhancing the reliability of the identification results. Finally，the proposed method is applied to identify tight gas channel sandstones in a work area in southwestern China. Compared with traditional seismic attribute analysis and intelligent identification results relying on a single data type，the proposed method demonstrates higher efficiency and reliability，validating its applicability.

Intelligent seismic velocity inversion is currently a hot and challenging topic in seismic exploration research. Nevertheless, the complex structure of deep learning networks demands significant computing power from hardware devices, which restricts the application of the model in scenarios with large data volumes and high timeliness requirements. To address these practical issues, in this paper, the U-Net is improved based on the concepts of feature engineering and model lightweighting, and the inversion networks U-Net vG for GPU and U-Net vC for CPU are proposed. Firstly, the characteristics of the velocity inversion network are analyzed to deduce the lightweighting principles of convolutional neural networks. Subsequently, lightweight processing is conducted on the multi-scale module, attention gate module, and feature extraction module to obtain a lightweight convolutional neural network for velocity modeling, which reduces the network volume while maintaining prediction accuracy. Data test results demonstrate that the training process of the proposed network has lower requirements for high-performance hardware resources, and that the network enables efficient velocity inversion, possesses higher seismic velocity inversion accuracy, and exhibits superior noise resistance. It provides a new idea for solving the computing power bottleneck problem in seismic data inversion.