Deterministic model of earthquake clustering shows reduced stress drops for nearby aftershocks

Abstract

While a number of viable physical mechanisms have been offered to explain the temporal clustering of aftershocks, the spatial clustering of aftershocks, in particular the concentrated productivity of aftershocks very near the mainshock rupture area, has been difficult to reproduce with physical models. Here we present a new deterministic physical model capable of reproducing both the spatial and temporal clustering. We apply this new model to a longstanding puzzling question raised by ground motion observations, which suggest that nearby aftershocks show reduced ground motions relative to similar magnitude mainshocks. In the model, the physical basis for these observations is reduced stress drops for nearby aftershocks compared to similar magnitude mainshocks. These reduced stress drops are due to nearby aftershocks rerupturing incompletely healed parts of the fault which ruptured in the mainshock.

Low stress drops observed for aftershocks of the 2011 Mw 5.7 Prague, Oklahoma, earthquake

“…Spatially varying stress drops may reflect strong fault heterogeneity [e.g., Fehler and Phillips , ], which for induced events may be influenced by the injection of fluids into the subsurface. Alternatively, low stress drops may be attributed to nearby aftershocks that re‐rupture parts of the mainshock fault plane that may not have completely healed [e.g., Vidale et al , ; Shaw et al , ]. It is unclear if the variation we observe here is a reflection of actual physical properties of the fault or simply due to scatter in the measurement due to changing station configuration and/or errors in corner frequency estimates.…”

“…Abercrombie [2014] reported that stress drops along the San Andreas Fault at Parkfield decrease following the 2006 M6.0 mainshock and then rebound rapidly to previous levels. She attributes the temporal variability to short recurrence time that limits healing time and may reflect lower fault strength [e.g., Vidale et al, 1994;Shaw et al, 2015]. Alternately, the apparent temporal correlation we observe here could simply reflect lateral or depth variability in the fault strength and changing event locations.…”

Spatial and Temporal Variations in Earthquake Stress Drop on Gofar Transform Fault, East Pacific Rise: Implications for Fault Strength

“…Similarly, stress drops for aftershocks in the 2007 M W 6.2 rupture asperity also decreased, suggesting these earthquakes occurred in the region of strong ground motion from the M W 6.0 mainshock. Low stress drop aftershocks (< 0.5 MPa) may reflect the rupture of fault segments that have not fully healed from the large earthquake ground motion (Shaw et al, ). In contrast, high stress drop (> 1 MPa) aftershocks may reflect strong asperities in a damaged and otherwise weakened fault zone (e.g., Dreger et al, ) or reflect stress concentrations at the fault edges (Oth & Kaiser, ).…”

Aftershock Ground Motion Record Selection: A Novel Mainshock-Consistent Approach

“…Moreover, there is an ongoing debate in the literature regarding the frequency content of AS ground motions compared to MS ones [e.g. 17,18]. Nearby (to the MS rupture) ASs tend to exhibit lower stress drops compared to MS events with similar magnitude.…”