Evolution of induced earthquakes from dimensionless scaling

TL;DR

This paper analyzes the evolution of induced earthquakes using dimensionless scaling, revealing bimodal distributions and proposing a new parameter-free statistical forecasting method that models exponential growth in earthquake event counts.

Contribution

It introduces a novel dimensionless scaling approach to model induced earthquake growth and proposes a parameter-free forecasting method that avoids magnitude cut-offs.

Findings

Bimodal distribution of earthquake counts in the US and Groningen.

Exponential growth in induced earthquake events with a doubling time of 6.24 years.

A tabletop experiment illustrating shear stress and crack formation.

Abstract

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.