Master event and double-difference techniques were used to relocate mining-induced seismicity (MIS) at the Trail Mountain Mine, a longwall coal mine in central Utah. Travel-time data were collected by Arabasz et al. (2002) using a surface seismic network with stations at elevations both above and below mine level (because of the topography) and a single in-mine station. Arabasz et al. (2002) only used surface stations above mine level to determine locations. Using this network geometry, they were only able to constrain focal depths for 321 of 1829 events. In contrast, we use all stations, creating a 3D network. Hypocentral locations are improved by implementing a master event methodology to reduce the effects of uncertainties in the velocity structure, though the resulting locations do not correspond with known structures or stratigraphy. The mismatch between the locations and geology is likely due to fracturing of the rock mass by the mining process, thereby decreasing the seismic velocity near mined-out regions. Assuming a 10% velocity decrease places the MIS in the roof of the mine. A double-difference procedure is used to mimic a timevarying velocity structure. The time-varying velocity structure results in locations that approximate the dip of the coal seam. By using all available stations and allowing for a time-varying velocity structure, we find the MIS is located immediately above the coal seam and closely follows the position of the coalface. The epicenters align with the roads along the longwall panel, where stress concentrations are expected during mining.Online Material: Animations of the progression of seismicity along the longwall panel.

Summary Monitoring mining-induced seismicity (MIS) can help engineers understand the rock mass response to resource extraction. With a thorough understanding of ongoing geomechanical processes, engineers can operate mines, especially those mines with the propensity for rockbursting, more safely and efficiently. Unfortunately, processing MIS data usually requires significant effort from human analysts, which can result in substantial costs and time commitments. The problem is exacerbated for operations that produce copious amounts of MIS, such as mines with high-stress and/or extraction ratios. Recently, deep learning methods have shown the ability to significantly improve the quality of automated arrival-time picking on earthquake data recorded by regional seismic networks. However, relatively little has been published on applying these techniques to MIS. In this study, we compare the performance of a convolutional neural network (CNN) originally trained to pick arrival times on the Southern California Seismic Network (SCSN) to that of human analysts on coal-mine-related MIS. We perform comparisons on several coal-related MIS datasets recorded at various network scales, sampling rates, and mines. We find that the Southern-California-trained CNN does not perform well on any of our datasets without retraining. However, applying the concept of transfer learning, we retrain the SCSN model with relatively little MIS data after which the CNN performs nearly as well as a human analyst. When retrained with data from a single analyst, the analyst-CNN pick time residual variance is lower than the variance observed between human analysts. We also compare the retrained CNN to a simpler, optimized picking algorithm, which falls short of the CNN's performance. We conclude that CNNs can achieve a significant improvement in automated phase picking although some dataset-specific training will usually be required. Moreover, initializing training with weights found from other, even very different, datasets can greatly reduce the amount of training data required to achieve a given performance threshold.

Deformation and support conditions in underground mines are typically monitored through visual inspection and geotechnical instrumentation. However, the subjectivity of visual observation techniques can result in ambiguous or incomplete analyses with little quantifiable data. Monitoring displacements with conventional instrumentation can be expensive and time-consuming, and the information collected is typically limited to just a few locations. Moreover, conventional methods usually provide vector rather than tensor descriptions of geometry changes, the latter of which offer greater insight into rock movements and potential ground fall hazards. To address these issues, researchers from the National Institute for Occupational Safety and Health's Spokane Mining Research Division have developed and evaluated photogrammetric systems for ground control monitoring applications in underground mines. In cooperation with the Hecla Mining Company, photogrammetric surveys were conducted over a three-year period at the Lucky Friday mine in northern Idaho, United States of America, as underhand cut-and-fill mining methods were used to mine Ag-Pb-Zn ore in rockburst-prone ground conditions at depths approaching 2,100 metres. A photogrammetric system was developed for underground use at the mine that is not only mobile, rugged, and relatively inexpensive, but also capable of producing measurements comparable to conventional displacement-measuring instrumentation. This paper describes the components of the photogrammetric system, discusses the use of point cloud analyses from photogrammetric surveys to monitor bulk deformation in underground entries, and explains the advantages of full tensor descriptions of three-dimensional (3D) ground movement, particularly in regard to the interpretation of potential movement along fault intercepts. The practical use of photogrammetry to augment measurements from conventional instruments, such as crackmeters, is presented, as well as the use of photogrammetric data in conjunction with 3D visualisation software to synthesise and integrate complex information from diverse sources including geology, mining configuration, seismicity, and geotechnical instrumentation.

Over the past decade, ObsPy, a python framework for seismology (Krischer et al., 2015), has become an integral part of many seismology research workflows. ObsPy provides parsers for most seismological data formats, clients for accessing data-centers, common signal processing routines, and event, station, and waveform data models.

This document is in the public domain and may be freely copied or reprinted. Disclaimer Mention of any company or product does not constitute endorsement by the National Institute for Occupational Safety and Health (NIOSH). In addition, citations to Web sites external to NIOSH do not constitute NIOSH endorsement of the sponsoring organizations or their programs or products. Furthermore, NIOSH is not responsible for the content of these Web sites. All Web addresses referenced in this document were accessible as of the publication date.