Subducted seafloor relief stops rupture in South American great earthquakes

This was my first published paper, based on research undertaken during my Masters. It is published in Earth and Planetary Science Letters, and available for download the MMU e-space repository.

The initial observation for this work was that ‘great’ earthquakes, those which measured more than magnitude 8.0, tend to have end points in the same place. An earthquake end point is the limit to the earthquake fault plane movement, shaking will take place outside of this zone but it is most violent in the regions above the fault plate movement. The figure below shows the rupture zones and end points for great South American earthquakes.

Rupture zones and subducting topography in South America
Rupture zones and subducting topography in South America

An initial inspection of the plate margin (where the incoming Pacific Plate is subducted underneath the South American Plate) suggested that when there are underwater mountains on the Pacific Plate coming into the subduction zone (black blobs on the figure above), these tend to match up with the end points of earthquakes. The figure below shows how topographic features (underwater mountains and ridges) match up with earthquake locations.

Subducting topography and earthquake locations
Subducting topography and earthquake locations

To test whether this was a real relationship, or just coincidence, I designed a model that produced earthquakes along the subduction zone. The model had two versions. In version one, earthquakes were placed in line along the subduction zone (as tends to happen in these situations, one earthquake starts at the end point of a previous one) but their end points were not constrained by anything. In the second version, if the earthquake tried to rupture past an incoming topographic feature, the rupture was stopped there.

By applying statistical tests to the model, I showed that the endpoints of earthquakes were far more likely than random chance to be located where topography > 1000m was being subducted. The model also showed that subducting topography led to a reduction in background earthquakes. At earthquake endpoints not associated with topography, there was an increased amount of smaller earthquakes releasing the built-up stress, but this was not seen at earthquake endpoints near to subducting topography. Therefore, subduction of a high seamount or ridge makes the subduction zone earthquake activity decrease (it becomes ‘aseismic’) which prevents both great earthquakes getting past, and smaller earthquakes from occurring.