Multistation template matching to characterize frequency-magnitude distributions of induced seismicity in the Central and Eastern US

Abstract:

We analyze the frequency-magnitude distribution (FMD) of recent seismic sequences thought to be induced by wastewater injection and hydraulic fracturing in the Central and Eastern U.S. to investigate their physical origin and improve hazard estimates. Multistation template matching is utilized to increase the number of events analyzed by lowering the magnitude of detection. In cases where local deployments are available, we demonstrate that the FMD obtained through template matching using regional data are comparable to those obtained from traditional detection using the local deployment. Since deployments usually occur after seismicity has already been identified, catalogs constructed with regional data offer the advantage of providing a more complete history of the seismicity. We find two primary groups of FMDs for induced sequences: those that generally follow the Gutenberg-Richter power-law and those that generally do not. All of the induced sequences are typically characterized by swarm-like behavior, but the non-power-law FMDs are also characterized by a clustering of events at low magnitudes and particularly low aftershock productivity for a continental interior. Each of the observations in the non-power law FMD cases is predicted by numerical simulations of a seismogenic zone governed by a viscoelastic damage rheology with low effective viscosity in the fault zone. Such a reduction in effective viscosity is expected if fluid injection increases fluid pressures in the fault zone to the point that the fault zone begins to dilate.