Evolution of induced earthquakes from dimensionless scaling
Event counts are a powerful diagnostic for earthquake analysis. We report on a bimodal distribution of earthquake event counts in the U.S. since 2013. The new peak is about a magnitude M1 with a distribution very similar to that of induced earthquakes in Groningen, The Netherlands, characterized by exponential growth in event count since 2001. The latter shows a doubling time of 6.24 years with a relatively constant rate of land subsidence. We model its internal shear stresses as a function of dimensionless curvature and illustrate the resulting exponential growth in a tabletop crack formation experiment. Our study proposes a new method of parameter-free statistical forecasting of induced events, that circumvents the need for a magnitude cut-off in the Gutenberg-Richter relationship.