This paper utilizes prevalent deep learning techniques, such as Convolutional Neural Networks (CNNs) and Residual Neural Networks (ResNets), along with the well-established machine learning technique, Random Forest, to efficiently distinguish between common lithologies including coal, sandstone, limestone, and others. This approach is highly significant for resource extraction—such as coal, oil, natural gas, and groundwater—by streamlining the process and minimizing the need for the time-consuming manual interpretation of geophysical logging data. The natural gamma ray, density, and resistivity log data were collected from 22 wells in the mountainous region of Southern Sichuan, China, yielding approximately 70,000 samples for developing lithofacies prediction models. All the models achieved around 80% accuracy in classifying carbonaceous lithologies and up to 88% accuracy in predicting other lithologies. The trained models were applied to the logging data in the validation dataset, and the outputs were validated against geological core data, showing overall consistency, although variations in the classification results were observed across different wells. These findings suggest that deep learning techniques have the potential to develop a general model for effectively handling lithology classification with well logging data.

The transient electromagnetic method (TEM) is widely used for detecting low-resistivity areas ahead of tunnel. However, implementing the 2-D or 3-D inversion of whole-space geo-electric models is not feasible due to the narrow space within the underground tunnel and the nonuniqueness of TEM inversion. To solve this problem, we develop a fast inversion operator guided by deep learning (DL), which translates the time-domain TEM measurements directly into the spatial probability of water-filled anomalies’s position. Trained by synthetic data, our system shows impressive adaptability to predict water-filled anomalies when implementing different transmit currents, source waveforms, coil turn numbers, and abnormal body sizes. Compared to traditional 1-D tunnel TEM inversion, our system demonstrates less ambiguity, superior stability, applicability, noise resistance, and higher computational efficiency. The effectiveness of this method has been further confirmed by physical model experiments and field data. This inversion operator can support instantaneous TEM detection of low resistivity in the tunnel activities.

The probing depth of the transient electromagnetic method (TEM) refers to the depth range at which the underground conductivity changes can be effectively detected. It typically ranges from tens of meters to several kilometers and is influenced by factors such as instrument parameters and the conductivity of the subsurface structure. Rapid and accurate probing depth is useful for the selection of appropriate inversion parameters and improving survey accuracy. However, mainstream methods suffer from issues such as low computational precision, large uncertainties, or high computational requirements, making them unsuitable for processing massive airborne electromagnetic data. In this study, we propose a prediction model based on deep learning that can directly compute the probing depth from the TEM responses, and its effectiveness and accuracy are validated through synthetic models and field measurements. We compared the performance of classic deep learning models, including CNN, RESNET, and RNN, and found that RNN performed the best overall on both synthetic and field data. Furthermore, we apply this algorithm to deep learning-based ATEM inversion by constraining the one-dimensional resistivity model depths in the training set, to reduce the non-uniqueness of the inversion, accelerate the convergence, and improve its prediction accuracy.

Airwave interference presents a major source of noise in seismic exploration, posing significant challenges to the quality control of raw seismic data. With the increasing data volume in 3D seismic exploration, manual identification methods fall short of meeting the demands of high-density 3D seismic surveys. This study employs the YOLOv5 model, a widely used tool in object detection, to achieve rapid identification of airwave noise in seismic profiles. Initially, the model was pre-trained on the COCO dataset—a large-scale dataset designed for object detection—and subsequently fine-tuned using a training set specifically labeled for airwave noise data. The fine-tuned model achieved an accuracy and recall rate of approximately 85% on the test dataset, successfully identifying not only the presence of noise but also its location, confidence levels, and range. To evaluate the model’s effectiveness, we applied the YOLOv5 model trained on 2D data to seismic records from two regions: 2D seismic data from Ningqiang, Shanxi, and 3D seismic data from Xiushui, Sichuan. The overall prediction accuracy in both regions exceeded 90%, with the accuracy and recall rates for airwave noise surpassing 83% and 90%, respectively. The evaluation time for single-shot 3D seismic data (over 8000 traces) was less than 2 s, highlighting the model’s exceptional transferability, generalization ability, and efficiency. These results demonstrate that the YOLOv5 model is highly effective for detecting airwave noise in raw seismic data across different regions, marking the first successful attempt at computer recognition of airwaves in seismic exploration.