Landward thrusting in accretionary wedges: evidence for seafloor rupture?

Abstract:

The 2004 Sumatra and 2011 Japan earthquakes took the community by surprise because they ruptured frontal sections of megathrust thought to slip aseismically. Studying the deformation of accretionary prisms can help in characterizing the specific structures associated to frontal propagation and determining the mechanical properties leading to this behavior. Recent observations suggest a correlation between landward faults and frontal propagation of earthquakes along the Sumatra subduction zone. Large sections of landward thrusts are also observed along Cascadia known to have ruptured in 1700 with a M~9 generating a large tsunami.

In this study, we propose to investigate if specific frictional properties could lead to a landward sequence of thrusting with the limit analysis approach. We first show that such sequence requires very low effective friction along the megathrust with a rather high internal effective friction. We also show that landward sequence appears close to the extensional critical limit. We retrieve the megathrust effective friction for three wedges with different sediment incomes. For Cascadia, we find a maximal effective friction of 0.032. For northern and southern Sumatra, we find μ≤0.02 and μ≤ 0.08 respectively. This very low effective friction is probably due to lithostatic pore pressure.

This high pore pressure could either be a long-term property or due to dynamic effects such as thermal pressurization. The fact that landward vergence appears far from the compressional critical limit favors a dynamic effect. Indeed, a wedge would move away from this limit if material is added synchronously to the deformation or if it is suddenly submitted to a lower effective friction. In addition, the long-term high pore pressure could be due to a low permeability enhancing thermal pressurization and co-seismic slip along the frontal part of the megathrust.