Temperature rise, slip localization, and fault weakening as observed from the rock record

Abstract:

There is ample laboratory and theoretical evidence that faults experience significant frictional weakening during earthquake slip, and many weakening mechanisms are linked to temperature. We summarize observations of coseismic temperature rise in natural faults including: relationship with depth, slip zone thickness and frictional energy. Even when considering earthquakes of various sizes, temperature rise is limited at depths less than 5 km, reflected by the rarity of pseudotachylytes in the shallowest crust. More surprisingly, temperature rise is fairly constant below 5 km and slipping zones get slightly thicker with depth. We attribute this to higher ambient temperatures allowing faults to achieve the temperature rise needed for dynamic weakening without significant localization.

From the maximum temperature estimates and fault zone thickness, we can solve for average frictional energy. The frictional energy for almost all of the earthquakes in our dataset is below ~50 MJ/m², with the exception of the Pasagshak megathrust (105-228 MJ/m²), the only exhumed megathrust from seismogenic depths. We also compare frictional energy estimates to other parts of the earthquake budget, including fracture and radiated energy. Finally, we discuss the few key examples of delocalized slip during earthquakes and what important mechanisms we might be missing by focusing solely on temperature-driven weakening.