The past few decades have seen a number of very large earthquakes at subduction zones. Researchers now have an array of advanced technologies that provide insights into the processes of plate movement and crustal deformation. A review article recently published in Reviews of Geophysics pulled together observations from different locations worldwide to evaluate whether similar physical processes are active at different plate margins. The editors asked one of the authors to describe advances in our understanding and where additional research is still needed.
What are “megathrust zones” and what are the main processes that occur there?
A megathrust zone is a thin boundary layer between a tectonic plate that sinks into the Earth’s mantle and an overriding plate. The largest earthquakes and tsunamis are produced here. High friction in the shallow part of the megathrust zone effectively locks parts of the interface during decades to centuries. Ongoing plate motion slowly brings the shallow interface closer to failure, i.e., an earthquake. Other parts of the megathrust zone are mechanically weaker. They consequently attempt to creep at a rate that is required by plate tectonics, but are limited by being connected to the locked part of the interface.
What insights have been learned from recent megathrust earthquakes at different margins?
High magnitude earthquakes in Indonesia (2004), Chile (2010) and Japan (2011) were recorded by new networks utilizing Global Positioning System technology, which is capable of measuring ground displacements with millimeter accuracy. This complemented seismological observations of megathrust slip during these earthquakes. The crust turned out to deform significantly during and after these earthquakes. These observations indicated that slip on weak parts of the megathrust zone may be responsible, likely in combination with the more classical stress relaxation in the Earth’s mantle. In regions where megathrust earthquakes are anticipated, crustal deformation observations allowed researchers to identify parts of the megathrust zone that are currently locked. In our review article, we integrate these perspectives into a general framework for the earthquake cycle.
How have models been used to complement observations and better understand these processes?
Mechanical models are needed to tie the surface observations to their causative processes that take place from a few to hundreds of kilometers deep into the Earth, which is beyond what is directly accessible by drilling. Many of the published models focus on a single earthquake along a specific megathrust zone. We wondered what deep earth processes are common to these regions globally and built a unifying model to predict its surface expressions. Our model roughly reproduced the observed surface deformation, but it also became clear that some regional diversity would be required to match the data shortly after a major earthquake.
What have been some of the recent significant scientific advances in understanding plate boundaries?
Creep on weak parts of the megathrust zone is a very significant contributor to the surface measurements after an earthquake. Mantle relaxation is also relevant. We demonstrate that the surface deformation of these processes may give a biased impression of low friction on the megathrust zone. Creep on the megathrust zone downdip of a major earthquake may be responsible for observations that were puzzling thus far; in an overall context of convergence and compression, tension was observed in the overriding plate shortly after recent major earthquakes.
What are some of the unresolved questions where additional research or modeling is needed?
Marine geodesy is an exciting new field that aims to monitor deformation of the sea floor that already yielded important constraints on the deformation of the Japan megathrust. Measurements along various margins will tell whether all megathrusts are locked all the way up to the seafloor. A longstanding question is how observations on geological time scales of mountain building and deformation of the overriding plate are linked to the observations of active deformation. We think that the multi-earthquake cycle model that we present in this review article is a first step towards that goal.
—Rob Govers, Faculty of Geosciences, Utrecht University, Netherlands; email: [email protected]