Detection of mining-induced microseismicity through a deep convolutional neural network
Bibliographic record
Abstract
The underground extraction of mineral resources is often closely linked to induced microseismic events. The use of a seismic network to continuously monitor mining-induced seismicity to reduce risks and improve operational safety is common. For this monitoring to be effective, a comprehensive catalog of microseismic events, containing low-to high-magnitude events, is essential to evaluate the response of the rock mass to mining activities. However, detecting low-magnitude events based on manual picking or automated conventional approaches has been challenging in mining environments owing to the inherent noise level. Recent advancements in deep learning and data-driven methods, particularly Convolutional Neural Networks (CNNs) trained on extensive seismic datasets, have shown improved capabilities in automated event detection and arrival phase picking on seismic data recorded by regional seismic networks. In this study, we assessed the performance of PhaseNet, a deep learning arrival-time picking method, in detecting the P- and S-wave arrivals of mining-induced microseismic events at different noise levels. As access to high-quality, labeled microseismic datasets for such mining applications is rare, a realistic three-component synthetic dataset was generated using full-waveform modeling. This simulation accounted for the geological conditions and network geometry specific to a mine in Ontario, Canada. The mine, which integrates copper and nickel operations, experiences considerable mining-induced earthquakes annually, posing risks to miners and infrastructure. The simulation includes a variety of source mechanisms with different magnitudes and offers more than 270,000 labeled seismograms. The results from the PhaseNet-trained model, which utilized the simulated dataset, demonstrated its effectiveness in managing noisy waveforms. This capability allows the detection of low-magnitude events within the mine environment, which may be overlooked by traditional methods. Furthermore, the model shows high accuracy in picking both the P- and S-wave arrival times, achieving precision rates exceeding 0.9. Tests on real data were performed in three different scenarios. The first scenario involves training the model exclusively using real data. The second scenario combines synthetic and real data to retrain the model previously trained with synthetic data only. Finally, the third scenario focuses on retraining the pre-trained model using only synthetic data. All these trained models were used to evaluate the performance on the real test dataset. The results indicate that the model retrained with synthetic and real seismograms yielded the best arrival time predictions for the mine dataset.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".