Earthquake supercycle in subduction zones controlled by the width of the seismogenic zone

Abstract:

Supercycles describe a long-term cluster of differently-sized megathrust earthquakes that consist of partial ruptures leading up to the final complete failure of a subduction zone segment. We recognize that supercycles are suggested to occur in those subduction zones (Sumatra, Japan, S-Chile, N-Central Chile, Colombia-Ecuador) for which the estimated seismogenic zone downdip width is larger than average (111 km). Here we provide an explanation for this potential link between the seismogenic zone downdip width and supercycles.

We use a two-dimensional, continuum-based seismo-mechanical model, which was recently validated through a comparison against scaled analogue subduction experiments. The setup consists of an upper plate represented by a visco-elastic wedge, which is underthrusted by a rigid plate. The megathrust is simulated by a layer governed by slip-rate dependent friction, in which the velocity-weakening seismogenic zone extends over a certain width and transitions up-and downdip to velocity-strengthening regions.

In our simulations, the first megathrust events in a supercycle generally rupture only the outermost parts of the seismogenic zone. These sub-critical and pulse-like ruptures are stopped due to a large excess of strength over stress, which leads to a transfer of stresses towards the centre of the seismogenic zone. In addition to the continued tectonic loading, they thereby gradually reduce the strength excess so that the largest megathrust events finally rupture the entire seismogenic zone in a crack-like manner and release most of the accumulated stress. A greater downdip width increases the average strength excess and thus favours supercycles over ordinary cycles of only similarly sized complete ruptures.

We conclude that such stress evolution along the dip of a wide seismogenic zone is the simplest mechanism governing supercycles. Additional a priori complexities, like previously suggested structural or frictional heterogeneities are not required to generate supercycles, although they are expected to add to our explanation. Our results imply that larger than thus far observed earthquakes could occur as part of a supercycle in subduction zones where downdip widths are larger than average such as in Alaska, Java, Antilles and Kamchatka.