Numerical modeling of subduction zones: thermo-mechanical stabilization as a function of overriding plate rheology and thickness

The absence of a forearc is a recurrent simplification in numerical subduction models. This because, to our knowledge, there are no previous studies that have systematically investigate the role of this structure on subduction systems. Despite its short length (166 ± 60 km), the forearc has a significant impact in the maintenance of a stable subduction. It has already been proposed that the serpentinization of this region, by percolating fluids from the sinking slab, reduces the effective mechanical strength of the plate coupling zone interface, allowing the one-sided asymmetric subduction observable in nature. Moreover, the forearc could be the key stabilization mechanism in intra-oceanic subduction settings. In this scenarios, the oceanic overriding plate (OP) could be in a thermal state such that would also be negative buoyant. The ubiquitous presence of forearcs in all-active intra-oceanic subductions suggests that a weak interface alone could not be enough to prevent the OP to sink. Adding a positive buoyant forearc  to the tip of the OP could provide the counterforce required to prevent the OP to sink, and eventually, double-sided subduction setting. There are studies that already implement a forearc structure in their numerical models. However, since its dynamic influence has not been study yet, we can not predict its impact and/or ascribe a specific dynamic behaviour of the system to it. 

In this work we investigate the role of the forearc and its contribution to emergent features in subduction zones. We present a series of fully dynamic, buoyancy driven, thermo-mechanical numerical modelling experiments with a free surface carried out to gain insight on the dynamic role of the forearc.  We use the Underwolrd numerical code to perform a parametrization to geometric and rheologic parameters of this structure, namely the thickness (age of the OP), length and density. We consider a forearc that encompasses the arc (25 to 250 km wide) as well. We kept all physical properties of the subducting plate  constant throughout all models. Therefore, we are able to ascribe all dynamic changes solely to variations of the forearc properties. We test different forearc compositions based on its density, ranging between 2700 and 3300  kg.m⁻³, for 200  kg.m⁻³ intervals, mimicking a full granitic continental and an basaltic oceanic forearc, respectively. For all densities, we also test several possible lengths, for 130 km and for 200 to 470 km, for intervals of 90 km. Additionally, we test all possible density-length combinations for five different OPs, in terms of age, ranging between 20 and 100 Myr, for 20 Myr intervals. 

We expect a higher accommodation of strain in the tip of the OP in models where the forearc is implemented. The presence of this structure could favor slab roll-forward before this reaches the 660 km discontinuity, enhance subduction velocities and generate a more pronounced orogenic topography. This features would be enhanced with the decrease of density and thickness and  the increase of length of the forearc.