Applied Geophysics最新文献

Many traditional denoising methods, such as Gaussian filtering, tend to blur and lose details or edge information while reducing noise. The stationary wavelet packet transform is a multi-scale and multi-band analysis tool. Compared with the stationary wavelet transform, it can suppress high-frequency noise while preserving more edge details. Deep learning has significantly progressed in denoising applications. DnCNN, a residual network; FFDNet, an efficient, flexible network; U-NET, a codec network; and GAN, a generative adversative network, have better denoising effects than BM3D, the most popular conventional denoising method. Therefore, SWP_hFFDNet, a random noise attenuation network based on the stationary wavelet packet transform (SWPT) and modified FFDNet, is proposed. This network combines the advantages of SWPT, Huber norm, and FFDNet. In addition, it has three characteristics: First, SWPT is an effective feature-extraction tool that can obtain low- and high-frequency features of different scales and frequency bands. Second, because the noise level map is the input of the network, the noise removal performance of different noise levels can be improved. Third, the Huber norm can reduce the sensitivity of the network to abnormal data and enhance its robustness. The network is trained using the Adam algorithm and the BSD500 dataset, which is augmented, noised, and decomposed by SWPT. Experimental and actual data processing results show that the denoising effect of the proposed method is almost the same as those of BM3D, DnCNN, and FFDNet networks for low noise. However, for high noise, the proposed method is superior to the aforementioned networks.

The deep learning algorithm, which has been increasingly applied in the field of petroleum geophysical prospecting, has achieved good results in improving efficiency and accuracy based on test applications. To play a greater role in actual production, these algorithm modules must be integrated into software systems and used more often in actual production projects. Deep learning frameworks, such as TensorFlow and PyTorch, basically take Python as the core architecture, while the application program mainly uses Java, C#, and other programming languages. During integration, the seismic data read by the Java and C# data interfaces must be transferred to the Python main program module. The data exchange methods between Java, C#, and Python include shared memory, shared directory, and so on. However, these methods have the disadvantages of low transmission efficiency and unsuitability for asynchronous networks. Considering the large volume of seismic data and the need for network support for deep learning, this paper proposes a method of transmitting seismic data based on Socket. By maximizing Socket’s cross-network and efficient longdistance transmission, this approach solves the problem of inefficient transmission of underlying data while integrating the deep learning algorithm module into a software system. Furthermore, the actual production application shows that this method effectively solves the shortage of data transmission in shared memory, shared directory, and other modes while simultaneously improving the transmission efficiency of massive seismic data across modules at the bottom of the software.

Qingshankou shale (Gulong area, China) exhibits strong acoustic anisotropy characteristics, posing significant challenges to its exploration and development. In this study, the five full elastic constants and multipole response law of the Qingshankou shale were studied using experimental measurements. Analyses show that the anisotropy parameters ϵ and γ in the study region are greater than 0.4, whereas the anisotropy parameter δ is smaller, generally 0.1. Numerical simulations show that the longitudinal and transverse wave velocities of these strong anisotropic rocks vary significantly with inclination angle, and significant differences in group velocity and phase velocity are also present. Acoustic logging measures the group velocity in dipped boreholes; this differs from the phase velocity to some extent. As the dip angle increases, the longitudinal and SH wave velocities increase accordingly, while the qSV-wave velocity initially increases and then decreases, reaching its maximum value at a dip of approximately 40°. These results provide an effective guide for the correction and modeling of acoustic logging time differences in the region.

Local strong seismic activity shows the potential to closely follow a renewal process, which is inconsistent with the overall seismic activity that aligns with the Poisson process. Given that existing methods for synthesizing stochastic seismic event sets cannot control local seismic activity, a method based on Monte Carlo simulations has been developed for synthesizing random seismic event sets where local strong earthquakes satisfy the renewal process. This method can synthesize seismic activities in a statistical area where the overall activity conforms to the Poisson process and the major seismic activities in local potential sources or faults follow the renewal process. This paper presents long- and short-scale approaches. The long-scale earthquake catalogs are suitable for reflecting the sequential characteristics of seismic activities. Meanwhile, the short-scale catalogs focus on the impacts of specific earthquake events within a group for a detailed understanding of hazards under certain conditions, making them suitable for studies on specific earthquake sequences and geological areas or situations requiring high temporal resolution. In the applications of short-scale sequences, we find that the equivalent occurrence rate method may overestimate the seismic hazard. This synthesis method for earthquake catalogs can simulate realistic seismic activities, thereby enhancing the accuracy of hazard analysis results and is suitable for seismic hazard analysis and earthquake insurance rate setting.

With the rapid development of human society and the surge of engineering construction, the disturbance of the geological environment has grown substantial. The influence of extreme climatic factors and the frequency of geological disasters have increased. Landslide is an important geological disaster. The foundational research in geological disaster detection or monitoring involves studying the physical phenomena associated with the occurrence of landslides, analyzing their signal characteristics, and indirectly determining their development and changes. This scientific area is worthy of in-depth exploration. However, research on the key design and specific implementation of electromagnetic instruments tailored for landslide detection remains lacking. This study is based on the interdisciplinary theoretical system of electronic circuits, embedded computer science, digital signal processing technology, geophysics, and engineering geology. It independently develops a functional prototype of an electromagnetic detector for landslide detection. This design combines the current level of development of electronic disciplines and applies the embedded multicore Da Vinci industrial-grade computer system in the field of earth detection instruments. This study then designs and implements a landslide detection system with independent intellectual property rights.

The modeling and inversion of the large-scale gravity field is the basis for exploring the density structure and geodynamics of the deep Earth. For this reason, the realization of fast, large-scale gravity modeling has been a hot topic in recent years. By integrating horizontal adaptive subdivision and the radial extension technique, this paper investigates the combination of computational technology and strategy with the goal of enhancing the accuracy and efficiency of large-scale gravity modeling. This study also employs the computational strategy that combines mixed-resolution digital elevation models with the above techniques. This strategy significantly improves efficiency without compromising accuracy. The computational technology and strategy proposed in this paper provide a novel approach for facilitating high-accuracy gravity modeling on a large scale.

Since the beginning of the 21^st century, major earthquakes have frequently occurred worldwide. To explore the impact of astronomical factors on earthquakes, in this study, the statistical analysis method of correlation is used to systematically analyze the effects of astronomical factors, such as solar activity, Earth’s rotation, lunar declination angle, celestial tidal force, and other phenomena on M ≥ 8 global earthquakes at the beginning of the 21st century. With regard to solar activity, this study focuses on the analysis of the 11-year and century cycles of solar activity. The causal relationship of the Earth’s rotation is not obvious in this work and previous works; in contrast, the valley period of the solar activity century cycle may be an important astronomical factor leading to the frequent occurrence of global earthquakes at the beginning of the 21^st century. This topic warrants further study.

The numerical dispersion phenomenon in the finite-difference forward modeling simulations of the wave equation significantly affects the imaging accuracy in acoustic reflection logging. This issue is particularly pronounced in the reverse time migration (RTM) method used for shear-wave (S-wave) logging imaging. This not only affects imaging accuracy but also introduces ambiguities in the interpretation of logging results. To address this challenge, this study proposes the use of a least-squares difference coefficient optimization algorithm aiming to suppress the numerical dispersion phenomenon in the RTM of S-wave reflection imaging logging. By optimizing the difference coefficients, the high-precision finite-difference algorithm serves as an effective operator for both forward and backward RTM processes. This approach is instrumental in eliminating migration illusions, which are often caused by numerical dispersion. The effectiveness of this optimized algorithm is demonstrated through numerical results, which indicate that it can achieve more accurate forward imaging results across various conditions, including high- and low-velocity strata, and is effective in both large and small spatial grids. The results of processing real data demonstrate that numerical dispersion optimization effectively reduces migration artifacts and diminishes ambiguities in logging interpretations. This optimization offers crucial technical support to the RTM method, enhancing its capability for accurately modeling and imaging S-wave reflections.

The hydraulic fracturing of horizontal wells is a key stimulation technology for unconventional tight oil/gas reservoirs. Good knowledge of the near-well stress field of a horizontal well can be helpful for the hydraulic fracture design optimization of new wells and refrac design optimization of fractured wells. Azimuth and dip data derived from either focal mechanisms of hydraulic fracturing-induced microseismic events or fracture attributes of hydraulic fracture networks can be used for new-well stress field inversion. In this work, we present a novel stress inversion method integrating azimuth, dip, and rake data from the focal mechanisms of hydraulically induced microseismic events and fracture attributes of hydraulic fracture networks. For those stages having sufficient reliable microseismic focal mechanisms, strike, dip, and rake data derived from microseismic focal mechanisms are taken as input data for stress inversion. Meanwhile, for those stages that have no microseismic events or insufficient reliable microseismic focal mechanisms, azimuth and dip data derived from fracture attributes of prebuilt hydraulic fracture network are used for stress inversion, along with azimuth, dip, and rake data derived from other stages with sufficient reliable microseismic focal mechanisms. Thus, the near-well stress field of each stage can be inverted, regardless of whether or not it has hydraulically induced microseismic events. The new method has been applied in the field surface microseismic dataset during hydraulic fracture stimulation. The results reveal that the inverted near-well stress fields are consistent with the stress orientation derived from shear-wave splitting analysis for sonic logs. This finding demonstrates that the stress inversion method based on strike, dip, and rake data derived from microseismic focal mechanisms and fracture networks can correctly obtain the azimuths of maximum and minimum horizontal stress.

Boundary faults of the Daxing Uplift and Langgu–Dachang Depression are located in the southeastern region of the Beijing Plain and directly control the sedimentation, tectonic evolution, and strong seismic activity of the plain. The Sanhe–Pinggu earthquake of Ms 8.0 occurred in 1679, but the active tectonic deformation characteristics of the boundary have been rarely discussed. In this study, the active tectonic deformation characteristics of the Daxing Uplift and Langgu–Dachang Depression boundary rupture were investigated by collecting and analyzing the results of previous works, supplementing three shallow-seismic-exploration control lines at locations where the data are lacking, and carrying out borehole combined profile exploration and optically stimulated luminescence dating at local breakpoints. Results show that the Daxing Uplift and Langgu–Dachang Depression boundary faults constitute an active tectonic deformation zone with ∼50 km distance between Mafang and Niubaotun towns and then extends to both ends to form a deep and large fault that cuts through the earth’s crust. The activity of the Daxing Uplift eastern boundary fault may be divided into two sections near Anding town, with the early-to-middle Late Pleistocene gradually weakening in the northwest and the Holocene gradually weakening in the southwest. Moreover, the activity of the Xiadian fault may be divided into two sections near the Chaobai River: the Holocene gradually weakening in the northwest and the early-to-middle Late Pleistocene gradually weakening in the southwest. The boundary fault of the Daxing Uplift and Langgu–Dachang Depression has an ∼43 km seismic gap around Niubaotun town, which has a high risk of Ms 6.0–7.0 earthquakes. This investigation into the active tectonic deformation characteristics of the boundary fault of the Daxing Uplift and Langgu–Dachang Depression is crucial for analyzing the strong earthquake rupture behavior and the future risk of strong earthquakes in this area. It also contributes greatly to the study of the tectonic pattern evolution of the North China Plain and Beijing Plain.