Structural uncertainty in full-plate reconstructions as a way to account for lost intra-oceanic plate boundaries

The past decade has seen the rise of fully kinematic palaeogeographic models that explicitly define the evolution of both plate boundaries and tectonic plates. Included in these models are (possible) interpretations of spreading systems in extinct ocean basins. Typically, the primary constraint on controlling these synthetic mid-ocean ridges is ensuring that at known (i.e. preserved in the geological record) subduction zones there is convergence, and that at modelled mid-ocean ridges there is divergence. The most common way this is expressed in models is through a quasi-stable triple junction. While obviously subject to large inherent uncertainties, the advantage of modelling such ocean basins is that they can provide an internally consistent model of (tectonic) ocean evolution, tied to the underlaying palaeomagnetic and palaeotectonic framework. Here we explore this inherent uncertainty in such synthetic ocean basins, by introducing the concept of ‘structural uncertainty’ within a full-plate model. We describe structural uncertainty as the answer to the question, “how much oceanic-oceanic subduction (i.e. not preserved in the geological record) is required to balance the modelled synthetic spreading ridges?” While an initial inclination that models tending to ‘zero’ might be best, we entertain the possibility that there is a range of ‘lost’ subduction. To assess this hypothesis, we also interrogate whole-mantle convection models that produce self-consistent plate tectonics to determine the proportion of subduction around or adjacent to continents (representative of what might be preserved in the geological record), and subduction occurring within ocean basins (representative of what might be lost to the geological record).