Borehole radar, due to its high resolution and extensive radial detection capability, has become an important geophysical tool for detecting complex subsurface fractures. The key to evaluating fracture detectability lies in accurate fracture modeling and numerical simulation. To overcome the limitations of conventional fracture modeling approaches, including geometric oversimplification, inadequate representation of aperture and fracture surface correlation, and incomplete characterization of roughness, we propose a multi-factor three-dimensional (3D) rough fracture modeling method. This method integrates two-dimensional (2D) image reconstruction with the Weierstrass-Mandelbrot (W-M) fractal function, which enables a comprehensive description of fracture geometry, surface roughness, aperture, and correlation between the surfaces of a fracture. Based on the developed models, full-wave electromagnetic simulations of borehole radar are conducted using the finite-difference time-domain (FDTD) method, and the effects of fracture attitudes on radar responses are systematically investigated. The simulation results indicate that variations in dip angle and dip direction significantly influence the characteristics of the borehole radar signals. Fracture surface roughness is also found to introduce perturbations in echo details. Furthermore, the radar migration imaging results are more conducive to the evaluation of fracture attitude, as systematic simulation analysis demonstrates a high morphological consistency between the radar migration imaging results and the geometric projection of the fracture onto the Borehole-Fracture Coupling Plane (BFCP). This is further confirmed by the centroid offset distance and the Intersection over Union (IoU). In addition, the “dip direction ambiguity” in omnidirectional borehole radar detection is revealed, where fractures symmetric about the BFCP generate highly similar radar responses, thereby increasing the difficulty of interpretation. The presented fracture modeling method and observed response patterns from fractures support the accurate detection of complex fractures using borehole radar.

In electrical resistivity tomography (ERT), the anomaly effects of different electrode arrays vary depending on the geological model. The appropriate combination of different electrode arrays can optimize detection performance and enhance the reliability of interpretation results. However, traditional inversion methods, constrained by single-array data, sparse observations, and ill-posed problem-solving, often yield low-resolution or inaccurate results. To address the resolution challenges in ERT inversion, inspired by the outstanding fusion and nonlinear mapping capabilities of multi-modal deep learning (DL) image methods, we propose the super-resolution ERT fusion network (SRERTF-Net), which utilizes traditional inversion results of multi-array as the initial models, efficiently leveraging and integrating prior physical information to achieve multi-array data joint super-resolution inversion. In SRERTF-Net, different down-sampling paths are employed to process the inversion results of various electrode arrays, while Inception modules are introduced to enhance feature extraction. Additionally, dense connections are implemented both within and across paths to effectively integrate complementary information from different arrays, ensuring robust multi-modal feature fusion. Finally, we designed training samples that include randomly generated typical structural models and comprehensive complex models, in order to enhance the practicality and adaptability of the network. Experiments on synthetic and field measured data indicate that SRERTF-Net outperforms other methods in terms of resistivity accuracy, resolution, and background performance.

Denoising is a crucial step in ground penetrating radar (GPR) data processing. Conventional denoising algorithms for GPR typically require selecting optimal processing parameters, which can be challenging to achieve in practical applications, resulting in unsatisfactory processing outcomes. In recent years, in order to address the issue of low accuracy in conventional GPR denoising algorithms, denoising neural networks have been applied in the field of GPR. Although conventional denoising neural networks have shown improvements in signal-to-noise ratio (SNR) in some cases, their performance is often inadequate when facing real GPR data with complex random noise, due to the training methods of the networks. To address the challenges in denoising of GPR data, a two-stage denoising method based on deep learning (DL) has been proposed. Initially, conventional GPR data processing is conducted, followed by training a denoising network model using both the processed and unprocessed signals. Leveraging the powerful nonlinear fitting capability of convolutional neural networks (CNNs), an end-to-end mapping relationship is established to obtain the final denoising network model, completing the two-stage denoising process. Finally, this letter validates the proposed two-stage denoising method using synthetic and field data. The radar data obtained through this two-stage denoising method not only improve mean squared error (mse) by 0.17 compared to conventional methods but also increase peak SNR (PSNR) by 8.1. Furthermore, there is a significant enhancement in the integrity of the waveform and the recovery of weak signals.

Traditional cross-hole radar tomography (CRT) usually cannot obtain high-resolution imaging results due to the nonlinearity and multisolution of inversion. To cope with these challenges, we propose a scheme to achieve high-resolution CRT for complex slowness models using deep neural networks (DNNs). Given the inherent difficulty in generating complex geophysical models in batches, by series of processing some remote sensing images from the remote sensing scene classification dataset, we create a real slowness model dataset. Then, we utilize 2-D U-Net to directly construct the mapping relationship between the low-resolution slowness model from the traditional method and the real slowness model. The superiority of our scheme is verified by both synthetic data and measured data. Our scheme can significantly suppress the false anomaly of traditional CRT results and accurately reconstruct the underground target’s geometry, position, and slowness value, and it has excellent accuracy and robustness. In addition, the response data of the slowness model reconstructed by our scheme are closer to the field data.

Noise attenuation is crucial in ground-penetrating radar (GPR) data processing. In recent years, deep learning (DL) methods have shown excellent performance in GPR denoising tasks, but they typically focus only on recovering the target signal, which can lead to over-denoising. To enhance the generalizability and the practicality of denoising networks, we propose a strategy to generate random dielectric models from natural image datasets, which can quickly construct model datasets with low redundancy and reasonable distribution. To enhance the fidelity of GPR denoising, we leverage the powerful nonlinear fitting capabilities of convolutional neural networks (CNNs) and introduce a closed-loop denoising network framework for GPR. The framework consists of a denoising sub-network and a noise extraction sub-network, effectively achieving signal-noise separation in noised GPR data. Specifically, the denoising sub-network is used to recover weak reflection signals and initially remove noise, while the noise extraction sub-network is used to restore the true noise, mitigating the problem of over-denoising. A key innovation of our approach is the integration of bandpass filtering, which enhances the robustness of network training and supports effective weak signal recovery. This network framework forms a closed loop through the residual loss between the signal-noise separation results and the noised GPR data, the closed-loop structure is capable of further refining the signal and noise prediction results of the two subnetworks, thereby enhancing the numerical accuracy of the signal-to-noise separation results. Finally, the effectiveness of the GPR closed-loop denoising network is verified from multiple perspectives using both synthetic and field measured data. The results indicate that our proposed method is more competitive in GPR denoising tasks.

Directional borehole radar (DBR) is a powerful tool for constructing the 3-D morphology of subsurface geological bodies. Direction of arrival (DOA) estimation is a key step in 3-D imaging. However, the existing DOA methods are constrained by factors such as DBR aperture, and can only identify one signal source within a time window. Consequently, when faced with multiple targets, DBR encounters challenges in effectively distinguishing them, especially when their echoes nearly overlap. In addition, the presence of interference waves makes echo overlap more likely to occur. So we propose a 3-D imaging method based on echo separation to solve this problem. This method first separates echoes of different targets through echo separation methods such as correlation method, deconvolution method, τ-ptransformation method, and moving window method. Subsequently, the azimuth is obtained through the multiple signal classification (MUSIC) algorithm, while the depth and radial distance are obtained through the inverse bi-static boundary scattering transform (IBBST), finally achieving the 3-D imaging of multiple geological targets. The effectiveness of the proposed method is demonstrated through synthetic data of both simple and complex fracture models, and its feasibility in practical applications is demonstrated through field data examination. This method improves the detection capability of DBR, opening up new possibilities for accurately mapping subsurface geological features.

The traditional electrical resistivity tomography (ERT) inversion methods typically produce low-resolution imaging results due to the nonlinear and bulk effect of two-dimensional inversion. In this paper, we propose to directly establish the mapping from the geoelectric model of traditional inversion results (input) to the actual geoelectric models (output) through the fully convolutional networks (FCNs), inspired by the robust nonlinear mapping capabilities of deep learning methods. We designed an ERT resolution enhancement network (ERTReNet) based on the prevailing U-Net architecture, which can conduct end-to-end training and enhance the resolution of traditional inversion imaging results. This methodology has been tested on both synthetic and field measured data. Resolution has been improved, and the resistivity value of both target and geological background are closer to the synthetic model comparing to the tradition method. This work aids in improving the accuracy of subsurface target identification in ERT and serves as a guide for more precise ERT inversion in the future.

The emergence of directional borehole radar (DBR) has made it possible to obtain the 3-D morphology of fractures through a single borehole. The use of a uniform circular array (UCA) as the array receiving antenna is an important way to achieve directional detection; however, the direction of arrival (DOA) method and 3-D imaging method with DBR still have limitations in the imaging capability of rough fractures. We, therefore, propose a method of performing migration first and then DOA estimation next, combined with the moving window method (MWM), to achieve multitarget 3-D imaging, where the DOA estimation adopts the idea of applying the multiple signal classification (MUSIC) algorithm directly. We use the multigrid finite difference time domain (FDTD) method for numerical simulation to verify the effectiveness of the proposed 3-D imaging method. Subsequently, the response of DBR to rough fractures was studied, and the results showed that the roughness characteristics of fractures would significantly change the imaging results but also bring more fracture feature information; we can even detect the rough fracture under unfavorable dip angle. The research provides a theoretical basis for the detection of fractures under complex conditions and has practical application value in the future.

China’s first Mars probe Tianwen-1 has successfully landed on the southern Utopia Planitia of Mars on May 15, 2021. The Zhurong rover is first equipped with a full-polarimetric Mars Rover Penetrating Radar (FP-RoPeR) system, aiming to map the subsurface fine structure and to find the potential underground water ice. However, different from the previous water ice detection of orbital radar, the FP-RoPeR signals will be affected by the induced field rotation (IFR) if electromagnetic (EM) waves propagate through rough interfaces. Therefore, in this article, we assess the IFR effects from rough interfaces on the circular polarization ratio (CPR) response of FP-RoPeR data, which is a significant parameter for water ice detection. The theoretical computation and numerical validation indicate that the depth, the number of rough interfaces, and relative permittivity are three vital parameters that affect the IFR effects; depth plays a more important role than the other two for FP-RoPeR system. The CPR estimation result will be with greater error in the shallow region (0–1 m). The relative error in the region of depth greater than 1 m can be guaranteed to be under 10%.

Borehole radar is a powerful tool for detecting subsurface fractures. Fracture modeling and numerical simulation are essential means to study fracture detectability. In this article, we first propose a method to construct a single fracture model, which combines the Baecher disk model and random rough surface and includes features, such as roughness, pinch-out, and irregularity. As the precise description of the fracture raises the requirement for accuracy of the simulation algorithm, the 3-D conformal finite-difference time-domain (CFDTD) method is used in this work. Numerical simulations of a sphere and rough fractures show that the CFDTD has higher calculation accuracy than conventional FDTD. Then, we analyzed how the fractures with different roughness affect the electromagnetic wave response. It is found that as the fracture surface becomes rougher, the wave scattered by it becomes stronger, more fracture contour-related information is obtained, and the fracture morphology is recovered better. Combined with accurate fracture modeling and high-precision numerical simulation, the electromagnetic response characteristics of different forms of fractures are obtained, which provides a basis for the accurate detection and interpretation of fractures in the future.

Full-polarimetric Ground-penetrating radar (GPR) is considered as a promising sensor for detecting buried targets. However, the polarimetric decomposition technique plays a crucial role in identifying and classifying targets which are buried in the sand under the surface. The decomposition techniques of full-polarimetric Ground-penetrating radar includes four decomposition methods, namely: (1) Pauli decomposition method, (2) H-α decomposition method, (3) Freeman decomposition method and (4) polarimetric anisotropy analysis method .This paper mainly applys Freeman decomposition method to recognition of metal surface plate, dihedral and metal ball. The potential of polarimetric target decomposition techniques to metal surface plate, dihedral and metal ball characterization and classification is shown which provides valuable information.

Handheld ground-penetrating radar (GPR) system is one of a number of technologies that has been researched as a means of improving landmine detection efficiency. However, as the measurement points are random and data are irregular for the human operator, it is difficult to display subsurface visualization imaging. Also detection of buried landmines by GPR normally suffers from very strong clutter that will decrease the image quality. To solve the problem, a modified migration algorithm was proposed to process irregular GPR data, which has both the advantage of migration that can improve signal-clutter ratio and the advantage of interpolation that produces the grid data set for visualization. An application to field data acquired in Afghanistan shows clear landmine image in both vertical profile and horizontal slice.

Polarimetric GPR requires accurate calibration of channel imbalance and crosstalk not only in the amplitude term but also in the phase term. Currently, there have some calibration techniques. Though these techniques are very easy to perform, they provide less accurate calibration results for the crosstalk. To improve on the accuracy of calibration, we have developed a mathematical formulation to calibrate polarimetric GPR data. We measured several scattering matrices to obtain the necessary calibration parameters. The calibration technique was tested from measurements conducted on dihedral corner reflector.

Generally GPR transmits and receives radio waves with a single polarization using two parallel antennas. But it is possible to improve the GPR ability of discrimination and imaging of subsurface targets by analyzing the backscattered wave with a variety of polarizations. So we are developing a full-polarimetric GPR system, including PC, network analyzer, rectangular coordinates robot, switch driver, and polarimetric antenna array. Polarimetric antenna array is used to transmit and receive both co-polarimetric and cross-polarimetric signals. Currently polarimetric GPR do not execute precise calibration. But good calibration can improve the classification ability of subsurface targets. So we introduced the calibration technique into the polarimetric GPR, and derived a calibration formula.

We have conducted GPR survey at a site which was partly excavated and filled with highly contaminated soils. The electrical properties and TPH concentration of the core samples were measured in the laboratory. It is verified that an inverse relation between TPH concentration and relative dielectric constant, and a direct proportional correlation between TPH concentration and electrical resistivity. LNAPL contamination area is illustrated by GPR data which shows the decreased radar signal amplitude.

Migration is an important signal processing method that can improve signal-clutter ratio and reconstruct subsurface image. Diffraction stacking migration and Kirchhoff migration sum amplitudes along the migration trajectory, which generally is hyperbolic. But when the ground surface varies acutely, the migration trajectory is not hyperbolic. To computer the migration trajectory need the technique of ray tracing. We introduce a method of ray tracing based on 3D velocity model. Firstly, we build the 3D velocity model depending on the estimation of both ground surface topography and velocities. Then we compute the travel time between transmitter, receiver and each subsurface scattering point, and search the propagation ray depending on the Fermat's principle. The method is tested by an experiment data acquired by the stepped-frequency (SF) CMP antenna GPR system. The target is a metal ball that is buried under a sand mound. A nice result of ray tracing is shown in the case.

In order to understand the capability of monitoring ground water movement by ground penetrating radar (GPR), we carried out a field survey in Mongolia. We controlled water production at a ground water source area and the change of ground water level at about 5 m in depth could be detected by GPR. The change of the ground water level was observed more than 15 m in the radial distance from the pumping well. The change of the vertical water content was also evaluated by GPR.