Microseismic Monitoring in Underground Excavation Environments

Microseismic monitoring has emerged as a vital tool for mapping and assessing the evolution of rock mass behaviour during underground excavations. By recording small-magnitude seismic events induced by rock fracturing, blasting or stress redistribution, this technique provides real-time insights into fracture initiation, propagation and interaction with geological structures. Arrays of geophone sensors, often installed in boreholes or tunnel walls, capture seismic waveforms that are processed through time–frequency analysis and event-classification workflows. Modern monitoring systems integrate numerical simulation and inversion algorithms to estimate source parameters such as location, magnitude and focal mechanism, enabling the identification of critical zones of damage or instability. Advances in machine-learning and deep-learning methods have further enhanced the automatic detection and classification of microseismic signals, allowing for the discrimination of rock fracture events from blasting noise or mechanical disturbances. The global uptake of microseismic monitoring spans deep mining, tunnelling and large-scale civil works, underpinning risk mitigation strategies for rock bursts, slope failures and support design. By delivering continuous, quantitative data on subsurface dynamics, microseismic monitoring is central to improving safety, optimising excavation sequences and validating predictive models in complex geological settings.

Research from Nature Portfolio

Recent studies have demonstrated the efficacy of capsule networks in the automatic classification of microseismic records with limited training samples. By segmenting each record into multiple time–frequency frames and extracting a set of temporal and spectral features, the proposed model achieved an accuracy exceeding 99%, markedly outperforming conventional convolutional neural networks and traditional machine-learning classifiers. This approach addresses the challenge of scarce labelled data in newly established mines, ensuring reliable discrimination of genuine microseismic events from noise and operator-induced vibrations. The hierarchical feature representation inherent to capsule networks preserves spatial and temporal dependencies, offering a robust framework for real-time event recognition in underground environments.

Microseismic Monitoring in Underground Excavation Environments publication trend

The graph below shows the total number of articles in microseismic monitoring in underground excavation environments across all publications each year (not limited to Nature Index journals).

Technical terms

Microseismic event: A low-magnitude seismic disturbance caused by small-scale rock fracturing or stress redistribution in a geological medium.

Geophone sensor: A device that converts ground motion into an electrical signal for seismic wave detection and recording.

Waveform classification: The process of categorising seismic signals based on temporal and spectral characteristics to distinguish event types.

Capsule network: A deep-learning architecture that captures hierarchical spatial and temporal relationships among features for robust pattern recognition.

Convolutional neural network: A class of deep-learning model designed to extract local features through convolutional layers, widely used for signal and image analysis.

References

1. Stability Analysis of Rock Slopes Containing Faults During Excavation using Microseismic Monitoring and Numerical Simulation. Journal of Intelligent Construction (2024).
2. Microseismic records classification using capsule network with limited training samples in underground mining. Scientific Reports (2020).
3. A hybrid recognition model of microseismic signals for underground mining based on CNN and LSTM networks. Geomatics Natural Hazards and Risk (2021).