Using ambient seismic noise correlations to image and monitor underground mines

Using ambient seismic noise to image and monitor elastic properties of the earth's crust has become increasingly popular over the last decade. In this talk I will show how this technique can be applied in a different setting: the seismic noise generated by human activity in an active underground mine. The seismic noise recorded by sensors in an underground mine is vastly different from the noise recorded in the earth's crust - it is localised, high-frequency, often monochromatic and unstable. However, I will show that if the seismic data is processed carefully, it can be used to construct good estimates of the seismic Green's function (or virtual source signals) between sensors. I will show that the direct arrivals in these virtual source signals can be used to invert for the velocity structure in an underground mine and that the coda part of the virtual source signals can be used to measure very small changes (~0.005 %) in the seismic velocities that are induced by known stress changes.