Törnman, W & Martinsson, J 2020, 'Reliable automatic processing of seismic events: solving the Swiss cheese problem', in J Wesseloo (ed.), UMT 2020: Proceedings of the Second International Conference on Underground Mining Technology, Australian Centre for Geomechanics, Perth, pp. 155-172, https://doi.org/10.36487/ACG_repo/2035_04
(https://papers.acg.uwa.edu.au/p/2035_04_Martinsson/)
Abstract: BEMIS (Bayesian estimation of mining-induced seismicity) is a fully automatic, near real-time, robust and 
self-learning seismic processing solution for mining-induced seismic events. A prototype solution is tested in parallel with IMS’s routine manual processing in LKAB’s underground mines in Malmberget and Kiruna, providing four times more accurate earthquake locations based on 290 known blasts, 40 times faster processing time that scales with computer power, and the ability to detect and locate up to six times more events given the same input data. In addition to a fully automatic system, BEMIS provides a variety of unique functions such as quality control of all results, self-learning adaptation and calibrations, tomography, and prediction models of future seismicity. In this paper, we summarise the results from different investigations throughout time and discuss the unique approach considered to obtain reliable auto-processing in a challenging, unknown and changing environment.

Keywords: mining-induced seismicity, automatic processing, statistical seismology, reliable seismic parameters