Precise relative magnitude and relative location estimates of low-yield industrial blasts in Pennsylvania

Low-yield explosion monitoring requires the use of nearby short-period observations, which exhibit sensitivities to geologic heterogeneity and have low signal-to-noise ratios compared to larger events. In this study, we analyze 843 seismic events using nearly 475,000 individual observations to compute precise relative locations of small (1 < M_L < 3) industrial explosions across Pennsylvania using local shear-wave and short-period, near-regional distance shear and surface-wave observations. We show that common-station, nearby event cross-correlation time-shift measurements reduce much of the complexity in wave propagation caused by regional geological heterogeneity. The resulting high-precision relative location estimates (with formal location uncertainties of tens of meters in some cases) allow us to image the time-dependent migration of the blast wall for several mines across Pennsylvania. In areas with two or more mines, the relative locations collapse a diffuse distribution of small-magnitude industrial events into discrete clusters associated with particular mining operations. We also use cross-correlation amplitudes to estimate more precise relative event magnitudes. We find that the relative magnitudes are generally consistent with the catalog magnitudes but improve the relationship between the reported amount of explosives used and event size for several mining operations throughout Pennsylvania. This work adds to existing demonstrations of how dense regional seismic networks are valuable for small-event monitoring and characterization, while also corroborating earlier works indicating the ability of cross-correlation methods to achieve precise relative location and magnituode estimates from local and regional observations of low-yield seismic sources.