Earthquake stress via event ratio levels: Application to the 2011 and 2016 Oklahoma seismic sequences

We develop a new methodology for determining earthquake stress drop and apparent stress values via spectral ratio asymptotic levels. With sufficient bandwidth, the stress ratio for a pair of events can be directly related to these low- and high-frequency levels. This avoids the need to assume a particular spectral model and derive stress drop from cubed corner frequency measures. The method can be applied to spectral ratios for any pair of closely related earthquakes and is particularly well suited for coda envelope methods that provide good azimuthally averaged, point-source measures. We apply the new method to the 2011 Prague and 2016 Pawnee earthquake sequences in Oklahoma. The sequences show stress scaling with size and depth, with the largest events having apparent stress levels near 1MPa and smaller and/or shallower events having systematically lower stress values. Plain Language Summary The paper develops a new method to measure the stress that drives an earthquake. A fundamental question in earthquake science is whether this driving stress is similar for all earthquakes with random variations or if there are systematic changes with earthquake size and depth. The new method compares the seismic signal envelopes of two nearby events to cancel out the complex changes the Earth imparts to the signal as the waves propagate from the source to the seismic recording stations. The method was applied to recent earthquakes in Oklahoma, including 16 from the 2011 Prague sequence and 6 from the 2016 Pawnee sequence. The paper finds there do appear to be systematic increases in stress with size for these earthquakes, as well as systematic increases in stress with depth.