Self-Organising Map Analysis of Seismicity Associated with Mining at Mount Charlotte Mine

Authors: Mikula, PA; Fraser, SJ; Lee, MF Download Paper

DOI https://doi.org/10.36487/ACG_repo/808_103

Cite As:
Mikula, PA, Fraser, SJ & Lee, MF 2008, 'Self-Organising Map Analysis of Seismicity Associated with Mining at Mount Charlotte Mine', in Y Potvin, J Carter, A Dyskin & R Jeffrey (eds), SHIRMS 2008: Proceedings of the First Southern Hemisphere International Rock Mechanics Symposium, Australian Centre for Geomechanics, Perth, pp. 653-666, https://doi.org/10.36487/ACG_repo/808_103



Abstract:
Seismicity data from the mining of the ROB5 stope at Mount Charlotte mine is analysed using an approach known as self organising maps (SOM). SOM is an established technique that can reveal relationships between variables in complex and diverse data sets in an objective fashion. However, the mining industry has generally not recognised the potential of the technique. SOM is unsupervised – no prior knowledge is required as to the nature or number of groupings within the data set. The SOM approach is complementary to the MS-RAP spatial clustering and seismicity analysis package used at Mount Charlotte mine. SOM can form clusters based on spatial position, but its particular strength is its ability to cluster on the basis of any set of user-defined seismic event attributes, not just on location. This paper describes SOM clustering based on location, magnitude and the time interval between successive events. For the Mount Charlotte analyses, the absolute time of the seismic events was not revealed to the SOM. Yet, unassisted, SOM identified a clear change in seismic behaviour in June 2007 which correlated with the last major seismic event affecting the ROB5 destressing sequence. It also identified post-blast trends that may assist forecasting of increasing or decreasing seismicity. SOM also identified trends suggesting that specific lengths of seismic quiescence preceded higher magnitude events around the ROB5 stope.

References: