Directional performance of an algorithm used to locate microseismic events in underground mines

Locating local, low-intensity earthquakes, also referred to as microseismic events, in the vicinity of an underground mine is an important problem for both mine safety and mine planning. A microseismic event produces a finite time-duration acoustic signal that may have a small time-bandwidth product and passively locating such an event is a challenging task. Here a localization algorithm is described that hypothesizes a source location and aligns the sensor signals by removing the propagation delay for that hypothesized location. The hypothesized source location with the maximum energy, calculated from the sum of the aligned signals, is the estimate of the source position. This paper extends the algorithm to include signal weighting. An expression for the variance of location error is presented for the weighted algorithm and compared with the unweighted variance. A unique aspect of the variance expression is that it provides the performance in a direction specified by a three dimensional unit vector, eƒ, which is very useful for applications that restrict operations to tunnel structures, such as mining seismic and in-building acoustics.