Numerical models of the magmatic processes induced by slab breakoff

Overview & Synopsis

After the onset of continental collision, magmatism often persists for tens of millions of years, albeit with a different composition, in reduced volumes, and with a more episodic nature and more widespread spatial distribution, compared to normal arc magmatism. Kinematic modelling studies have suggested that slab breakoff can account for this post-collisional magmatism through the formation of a slab window and subsequent heating of the overriding plate and decompression melting of upwelling asthenosphere, particularly if breakoff occurs at depths shallower than the overriding plate. To constrain the nature of any melting and the geodynamic conditions required, we numerically model the collision of two continental plates following a period of oceanic subduction. A thermodynamic database is used to determine the (de)hydration reactions and occurrence of melt throughout this process. We investigate melting conditions within a parameter space designed to generate a wide range of breakoff depths, timings and collisional styles. Under most circumstances, slab breakoff occurs deeper than the depth extent of the overriding plate; too deep to generate any decompressional melting of dry upwelling asthenosphere or thermal perturbation within the overriding plate. Even if slab breakoff is very shallow, the hot mantle inflow into the slab window is not sustained long enough to sufficiently heat the hydrated overriding plate to cause significant magmatism. Instead, for relatively fast, shallow breakoff we observe melting of asthenosphere above the detached slab through the release of water from the tip of the heating detached slab. Melting of the subducted continental crust during necking and breakoff is a more common feature and may be a more reliable indicator of the occurrence of breakoff. We suggest that magmatism from slab breakoff alone is unable to explain several of the characteristics of post-collisional magmatism, and that additional geodynamical processes need to be considered when interpreting magmatic observations.