Stress Drop Variation of Deep‐Focus Earthquakes Based on Empirical Green's Functions

Abstract

We analyze source characteristics of global, deep-focus (>350 km) earthquakes with moment magnitudes (Mw) larger than 6.0-8.2 using teleseismic P-wave and S-wave spectra and an empirical Green's functions approach. We estimate the corner frequency assuming Brune's source model and calculate stress drops assuming a circular crack model. Based on P-wave and S-wave spectra, the one standard deviation ranges are 3.5-369.8 and 8.2-328.9 MPa, respectively. Based on the P-wave analysis, the median of our stress drop estimates is about a factor of 10 higher than the median stress drop of shallow earthquakes with the same magnitude estimated by Allmann and Shearer (2009, https://doi.org/10.1029/ 2008JB005821). This suggests that, on average, the shear stress of deep faults in the mantle transition zone is an order of magnitude higher than the shear stress of faults in the crust. The wide range of stress drops implies coexistence of multiple physical mechanisms. Plain Language SummaryThe change of shear stress (i.e., stress drop) during an earthquake is thought to be larger for deeper earthquakes than shallow earthquakes because of higher overburden pressure. However, the observational evidence for stress drop dependence on depth is still inconclusive. We estimate stress drops of earthquakes deeper than 400 km from recorded ground motion spectra. We find that the median stress drop of deep earthquakes is about one order of magnitude higher than the stress drop of shallow (<50 km) earthquakes. This implies that the shear stress of deep faults is moderately higher than of faults in the crust. The wide range of our stress drop estimates suggests that various mechanisms producing deep earthquakes coexist.

Well Proximity Governing Stress Drop Variation and Seismic Attenuation Associated With Hydraulic Fracturing Induced Earthquakes

“…The corrected stress drop values would still be proportionally lower with less discrepancy between values derived from P and S phases, but the scaling remains the same (Figure S6). While the Brune model derived from a flat circular shear crack model (Brune, 1970, 1971) would be insufficient in describing the source parameters of earthquakes with nonnegligible non‐DC components and thus potentially affects the corner frequency estimations (e.g., Liu et al, 2020), we note that recent work by Roth et al (2020) and references therein point to induced earthquake in the region being consistent with shear reactivation of existing faults correlated with injection activity.…”

Stress Drops of Intermediate‐Depth and Deep Earthquakes in the Tonga Slab

“…However, this approach suffers a strong trade‐off between the source and path terms. On the other hand, the most commonly used group of methods computes empirical Green's functions (eGFs) by modeling spectral ratios of P, S, or S coda waves (e.g., Abercrombie, 2014; Abercrombie et al., 2017; Huang et al., 2016; Liu et al., 2020; Nakajima et al., 2013; Ruhl et al., 2017). However, the unusually high attenuation in the Tonga mantle wedge and high noise level in our data from temporary island‐based stations and OBSs make this approach unpractical.…”

Across‐Slab Propagation and Low Stress Drops of Deep Earthquakes in the Kuril Subduction Zone

“…However, the detection and unambiguous description of deep earthquake ruptures are challenging because they occur far from seismic stations. A range of different seismic methods have previously been applied to determine the source of deep‐focus earthquakes; for example, durations can be determined from the stacking of high‐frequency P‐waves (Houston et al., 1998; Poli et al., 2016), source time functions (Campus & Das, 2000), or corner frequency approaches (Liu et al., 2020; Poli & Prieto, 2014). Similarly, detailed studies of the geometry, spatial dimensions and complex slip histories combine multiple methods, often utilizing high‐frequency data.…”

“…Each event is represented by a different symbol (A: plus; B: circle; C: diamond; D: star; E: triangle; F: cross; G: octagon; H: square; see the corresponding source parameters in TableS1in the Supporting Information S1). Dark and light blue circles are stress drop estimates from andLiu et al (2020), respectively. Orange and red symbols correspond to solutions obtained assuming sub-vertical and sub-horizontal fault planes, respectively.…”

“…In addition, whilst some studies report a depth dependence in source properties which may be linked to a change in mechanism between intermediate-and deep-focus earthquakes (Liu et al, 2020;Vidale & Houston, 1993), others have shown no such dependency (Campus & Das, 2000) or found that the apparent increase of stress drop with depth can be explained by the increase of rigidity (Vallée, 2013). To further explore the behavior of deep earthquakes, in this study, we use a second-degree moments technique (Backus, 1977a(Backus, , 1977bClévédé et al, 2004;McGuire et al, 2001) to find robust characteristic values of source parameters of deep events in the Kuril subduction zone.…”