Earthquakes are a common hazard that threatens human safety and results in huge financial losses. The movement of active faults has been recognized as the main cause of seismicity. The fracturing and acoustic emission characteristics of active fault zone under normal-slip faulting can provide insights for understanding the formation mechanism of earthquakes, but are currently not well understood. This study attempts to transcend the aforementioned limitation via the discrete element method model. The cementation degree of fault rock had a profound influence on the fracture initiation-propagation-coalescence patterns. Fractures were first concentrated in the fault footwall when the fault zone was cemented. And, the uncemented fault zone was extremely vulnerable to fracture, where several parallel fracture networks occurred in this scenario. The fault footwall was absolutely ruptured by the normal-slip faulting for both cemented and uncemented fault zones. Moreover, the AE events evolution of the cemented fault zone was consistent with the changes of fracture propagation velocity, but a different phenomenon was seen for the uncemented fault zone. The maximum value of AE energy occurred randomly, which did not follow the peak fracture propagation velocity. The stress release and drop was evident in the cemented fault zone under the normal-slip faulting. This phenomenon became insignificant with the increasing distance to the fault footwall. However, the unencumbered fault zone did not experience a sharp stress escalation and drop. This is likely to be caused by the creep-slip of the active fault. Based on these findings, the numerical framework established in this study provides a basis for microseismic monitoring of active faults.
Aeromagnetic exploration is a cost-effective geophysical method that is not constrained by terrain and has become an effective tool for geological prospecting. However, aeromagnetic anomalies are still delineated mainly by human interpretation, with computer-assisted interpretation playing a supplementary role. Developing new methods for anomaly delineation and improving delineation efficiency therefore remain important challenges. Aeromagnetic anomalies are potential-field superpositions generated by geological bodies at different depths and scales: high-frequency signals mainly reflect shallow, small-scale magnetic bodies, whereas low-frequency signals mainly reflect deep, large-scale magnetic bodies. This study introduces the synchrosqueezed wavelet transform and proposes a method for identifying aeromagnetic anomaly boundaries. By selecting frequency slices, the relative burial depth and boundary distribution of magnetic bodies can be qualitatively evaluated. Model experiments and tests on real data show that the method can identify and delineate the boundaries of anomalies generated by geological bodies, providing theoretical significance and practical guidance for subsequent lithological identification and mineral-prospecting prediction.
Deep metal mines are subject to multiple interference sources, including high in-situ stress, strong mining-induced disturbances, and complex underground electromechanical noise. As a result, microseismic monitoring signals generally exhibit low signal-to-noise ratio (SNR) values, and useful waveform features are often buried in noise, severely limiting the accuracy and reliability of source location, focal-mechanism inversion, and rock-mass stability analysis. Although mathematical-transform, mode-decomposition, and data-driven deep-learning methods each have their merits, they generally suffer from strong dependence on manual parameter tuning, low computational efficiency, or high sensitivity to training-data quality, and thus struggle to reconstruct microseismic waveforms with high fidelity while effectively suppressing noise. We propose a denoising algorithm based on the heat-conduction anisotropic nonlinear partial differential equation (hereafter ANDPDE). By constructing a structure tensor and a diffusion tensor, the method adaptively controls the diffusion direction and strength. A gradient-magnitude-weighted diffusion coefficient enhances diffusion filtering in flat, weak-gradient regions to fully suppress noise, while suppressing diffusion in strong-gradient regions (first arrivals, peaks, and coda oscillations) to preserve signal details. A dynamic amplitude-threshold constraint further suppresses anomalous amplitude perturbations during diffusion, ensuring that the source dynamic characteristics of microseismic events are preserved. Numerical tests demonstrate stable denoising performance under varying noise levels and geological structures, and field data show that the algorithm raises the SNR of a strongly contaminated record from as low as −3.13 dB to 22.24 dB, markedly improving the detectability of weak microfracture events and the accuracy of first-arrival picking. Requiring neither training samples nor manual parameter tuning, the proposed physics-driven model is computationally stable and generalizable, and it can provide high-quality data for rockburst early warning, stress-evolution analysis, and fracture-mechanism studies in deep mines.
Gas-bearing reservoir prediction is of great significance for energy security and safe hydrocarbon production. Seismic inversion provides an effective means for characterizing reservoir gas-bearing properties, among which the P- to S-wave velocity ratio (Vp/Vs) is highly sensitive to natural gas occurrence. However, in many mature oilfields, the absence of S-wave logging data limits the accuracy of Vp/Vs estimation and weakens the reliability of gas-bearing reservoir identification. To address this problem, this study proposes an integrated gas-bearing reservoir prediction workflow that combines improved consolidation-parameterbased rock-physics modeling with pre-stack Bayesian inversion. First, S-wave velocity is predicted from conventional logging data using the improved rock-physics model, thereby providing reliable constraints for constructing the low-frequency initial model. Then, a Cauchy prior is introduced into the pre-stack threeparameter Bayesian inversion to enhance inversion resolution and improve the characterization of elastic parameters. Finally, gas-bearing reservoirs are identified based on low Vp/Vs anomalies. Field data applications demonstrate that the proposed method improves the prediction accuracy of Vp/Vs compared with conventional statistical rock-physics approaches. The low Vp/Vs anomalies obtained from inversion are consistent with gasbearing intervals indicated by well data, confirming the effectiveness of the proposed workflow. This method provides a practical and reliable technical approach for gas-bearing reservoir prediction in mature oilfields with incomplete S-wave logging information.
In recent years, the demand for shallow surface engineering exploration has increased, and the exploration method of high-frequency artificial signal source has shown great potential in the field of shallow exploration. Aiming at this direction, this paper proposes a three-dimensional numerical simulation method of electromagnetic field in space wavenumber mixed domain which can be applied to low frequency and radio frequency bands (10kHz–10MHz). The basic idea of this method is to transform the three-dimensional partial differential problem based on the Coulomb gauge vector potential and scalar potential secondary field into a one-dimensional partial differential problem with independent wavenumbers by doing two-dimensional Fourier transform along the horizontal direction. The parallelism between different wavenumbers is high. According to the propagation mechanism of up-going wave and down-going wave, the boundary conditions of the space wavenumber mixed domain equation are constructed. According to the characteristics of the equation, the continuity condition of the interface current density is constructed by the partial integration method, and then the weighted residual method is used to solve the boundary value problem. One of the advantages of this method is that the one-dimensional problem formed after Fourier transform has high parallelism. The second advantage is that the fixed bandwidth equations formed by the finite element method can be solved efficiently by the chasing method. The third advantage is that the depth direction is retained as a spatial domain, and the grid spacing can be sparsely processed, taking into account the calculation accuracy and calculation efficiency. The prism model is designed and compared with other three-dimensional algorithms to verify the correctness and efficiency of the proposed algorithm in the numerical simulation of low-frequency and high-frequency electromagnetic fields. The computational efficiency is more than one order of magnitude higher than that of the traditional finite element method, which provides a new idea for realizing large-scale and efficient three-dimensional simulation of RF electromagnetic fields.
