Search

Subducted sediments are thought to lubricate the subduction interface and promote faster plate speeds. However, global observations are not clear-cut on the relationship between the amount of sediments and plate motion. Sediments are also thought to influence slab dip, but variations in subduction geometry depend on multiple factors. Here we use 2D thermomechanical models to explore how sediments can influence subduction dynamics and geometry. We find that thick sediments can lead to slower subduction due to an increase of the megathrust shear stress as the accretionary wedge gets wider, and a decrease in slab pull as buoyant sediments are subducted. Our results also show that larger slab buoyancy and megathrust stress due to thick sediments increase the slab bending radius. This offers a new perspective on the role of sediments, suggesting that sediment buoyancy and wedge geometry also play an important role on large-scale subduction dynamics.

[1] The fate of a convergent continental margin is investigated. We perform a set of 2D numerical models to study how and why continental collision can evolve in different scenarios. Since the rheology of continental lithosphere has a major control on the dynamics of subduction, we explore a range of different lithosphere and lower crust viscosity values to understand their sensitivity on the possible scenarios. We find that with a rheologically layered crust both delamination and break-off are feasible. We identify three modes: (1) slab detachment, in which the lithospheric mantle and the crust are strongly coupled, subduction slows down and the slab eventually breaks; (2) delamination of the lithospheric mantle that separates from the crust and continue to subduct and (3) an intermediate mode where the lithospheric mantle and the crust remain partially coupled, resulting in an initial stage of delamination followed by the slow down and cessation of subduction.

[1] A century after the catastrophic event, the sources of the 1908 Messina, Southern Italy, earthquake and tsunami, which caused at least 60,000 deaths, remain uncertain. Through a simple backward ray-tracing method, we convert the tsunami travel-time data reported in a 100-years-old paper into distances and find that the sources of the earthquake and tsunami are different. Overturning a long-held assumption, reconsideration of the available tsunami, bathymetric, seismic, and seismological data indicates that the tsunami was generated by an underwater landslide.

[1] Combining tectonics, with seismological and geochemical data, we reconstruct the deformation history of the presently narrow Calabrian slab and the path of mantle circulation during the last 10 Ma. We show that during the slab deformation the mantle laterally flowed inside the back arc region permitting its retrograde motion and giving a seismological and volcanological record after 1–2 myr.