Towards a mechanical understanding of megathrust behaviour – MECHATHRUST

Prediction of subduction earthquakes mostly relies on interplate coupling models providing patterns of slip deficit between tectonic plates. These coupling patterns are interpreted in the framework of elasto-dynamics and rate-and-state friction laws.
However, this framework has been challenged by recent observations, showing that rheological and geometrical complexities have to be taken into account to understand the megathrust mechanics. To account for these complexities, dynamic thermo-mechanical simulations, incorporating equivalent rate-and-state friction laws have emerged. However, they were run at the whole subduction scale impeding a precise study of the forearc deformation to be compared with its short-term deformation.
To understand how these complexities affect the megathrust behaviour, we here propose to develop a visco-elasto-plastic thermo-mechanical model at forearc scale. This model will incorporate thermal evolution and its feedback on rock rheology, and will be able to drop the time steps at key geological moments of the simulations to capture slow motion of earthquake cycle conditioned by loading and structures that are self-consistent with long-term deformation. After the model development, we will explore the effect of different initial conditions (such as thermal state, sediment cover thickness, mineralogy transitions, accretion vs erosive prism, subducting features) on the short- and long-term deformation of the forearc.
We will then investigate selected profiles along well-instrumented subduction zones, to provide mechanical explanations to the observed short-term deformation. We will run simulations at the seismic cycle scale in these regions and then along less constrained subduction zones to improve their seismic hazard assessment.