Abstract
Where two lithospheric plates converge on the Earth, one of them disappears into the mantle. The dominant driving mechanism for plate motion is regarded to be `slab pull': the subducted plate, the slab, exerts a pulling force on the attached plate at the surface. However, what has been puzzling geodynamicists
... read more
since their discovery in the
seventies, is that shallow, almost horizontal subduction, as occurring below Peru or Central Chile, does not fit into this simple picture. This forms the motivation for this study. Several proposed mechanisms for shallow flat subduction are investigated by means of numerical modeling experiments using a thermo-chemical convection code.
Active trenchward motion of the overlying lithosphere can force young, and weak oceanic lithosphere to subduct without the presence of a significant slab and associated slab pull, and may lead to shallow flat subduction, in this case called lithospheric doubling. A weak, preexisting lithospheric fault system and a lubricating oceanic crust
are required to continue the convergence during flat subduction. A relatively strong mantle fixes the oceanic plate while being overridden, and prevents the slab from descending more steeply through this mantle. Lithospheric doubling pushes underlying mantle material through the major mantle phase transitions at 400 and 670 km depth. The resultant latent heat may increase the flat slab length significantly, up to 400 km.
An alternative explanation for the occurrence of shallow flat slabs is the subduction of oceanic plateaus, aseismic ridges or
seamount chains. These oceanic features all have a thickened crust (up to 35 km), which give rise to an increased buoyancy of the subducting plate. The amount of metastable basalt must be substantial to keep the plateau sufficiently buoyant. Bending the slab to the horizontal requires the slab strength to be limited to about 600 MPa, and its
age to about 60 Ma. The flat subduction below Peru is most likely to be caused by both described mechanisms, since South America has a 3-cm/yr westward absolute plate motion, and the subduction of the Nazca Ridge oceanic plateau occurs below Peru. Only a small range of the average upper mantle viscosity and basalt-to-eclogite reaction
rates can explain the observed slab geometry. The effect of the overriding lithosphere is estimated to be one to two times larger than the effect of the plateau subduction.
Finally, the viability of the present-day subduction processes in a younger, hotter Earth is quantified. Higher mantle temperatures resulted in more partial melting and a thicker oceanic crust, comparable to present-day oceanic plateaus. Flat buoyant subduction has therefore been suggested to have been more important in a younger Earth. Model results, however, suggest that already for a modest increase of the mantle temperature less than about 75 K, the mantle
becomes too weak to support flat subduction. Modern-type Benioff subduction remains a viable tectonic process for potential temperatures of at least 150 K higher than today, but for higher mantle temperatures, crustal delamination or slab detachment prohibits a continuous subduction process.
show less
