Abstract

A general correlation between maximum co-seismic fault slip and fault length has been well constrained by observations from numerous earthquakes occurring on many crustal faults. In spite of this global consistency, there are notable examples of co-seismic fault slip magnitudes that far exceed the expected maxima for the fault dimensions.  Two instances of extreme fault slip that occurred on short-to-moderate length upper-plate faults in subduction systems, where the megathrust lies at shallow depths are the 2016 Kaikoura (New Zealand) earthquake, and the 1855 Wairarapa (New Zealand) earthquake. In both cases, co-seismic fault slip was 5 to 10 times greater than expected from the rupture length - fault slip scaling relationships. The typical crustal fault model  is a fault with a brittle-to-ductile transition at depth; in this scenario, co-seismic slip is inhibited by viscous resistance from the deeper, ductile component of the fault.  In contrast, the tectonic characteristics of faults that experience extreme co-seismic slip, involve upper plate faults that truncate against the megathrust at seismogenic depths (i.e. are fully frictionally coupled over their entire depth extent). During a megathrust earthquake, the plate interface unlocks and upper-plate faults that extend to the ruptured plate interface transiently experience free-slip boundary conditions on both their upper (surface) and lower (megathrust) ends. As a result, such upper-plate faults can potentially experience full strain release (and therefore maximum slip), independent of their length. For appropriately oriented faults, this effect may be enhanced by co-seismic stress changes associated with the megathrust earthquake.  Geologic evidence of large displacement (and/or displacement rate) upper-plate faults in other subduction systems indicates this process may commonly occur. One example is a set of upper-plate faults along the Cascadia margin (near Newport, Oregon), that have strike-slip geologic slip rates, averaged over 10s of thousands of years, exceeding tens of mm/yr and approaching local plate convergence rates. These upper-plate Cascadia faults are also located where the plate interface is sufficiently shallow and seismogenic, indicating that these high-slip, upper-plate faults are likely frictionally locked over their entire depth range. In spite of the high overall slip rates of these upper-plate faults, because they are locked along their entire depth extent between earthquakes,  they may be unrecognized by inter-seismic geodetic observations.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call