Speculation on the Initiation of Back-Arc Extension

The current outlines of the Earth's tectonic boundaries are known to be transient features from the perspective of geologic time. The machinations of the planet have long conspired to both devour and create lithosphere at these boundaries. However, the mechanisms for some of the processes by which this is done are not necessarily perfectly understood. For example, though the anatomy of subduction zones are well characterized, it is unclear why some subduction zones result in compression within the over-riding plate while others generate extension. Several ideas have been posited for driving the back-arc spreading observed in these extensional examples including mantle diapirism, secondary mantle convection, upper plate retreat, and subduction induced tension in the over-riding plate. Some preliminary numerical modeling from this study may lend some support to the general idea of the last mechanism. In a sequence of simulations, subduction was initiated between idealized oceanic and continental plates and allowed to evolve over approximately 55 myrs. In these simulations, stable subduction zones formed after approximately 15 to 20 myrs and persisted until approximately 45 to 50 myrs at which point, the trench spontaneously retreated "oceanward", rifting the over-riding continental plate and "towing" a continental sliver along with it. The oceanic slab roll-back appears to be concurrent with the intrusion of a low viscosity asthenospheric "finger" into the contact region between the two plates. This low viscosity finger may have denatured the fluid dynamically derived "suction" force acting between the plates. This force likely partially supports the subduction zone geometry, and its mitigation may have allowed the trench to retreat from its position. However, the portion of the suction force acting between the oceanic plate and the crustal portion of the continental plate was not as affected, thus possibly allowing the retreating plate to continue to adhere to a fragment of the continental plate as it retreated, rifting the weakened plate in the process. Though these simulations involve several gross simplifications and potentially unrealistic parameters, they may provide some insight into how past and current instances of back-arc spreading as well as some instances of continental rifting may have originated.