A Parameter Space Exploration of Subduction Geometries

Subduction is the great devourer of our planet’s lithosphere, shaping much of the world’s surface as well as feeding one of the primary driving engines of tectonics. Though it is easy to consider most subduction systems to be more-or-less similar given their similar 1st order morphologies, the overall structures of subduction zones do vary quite significantly from one another and, frequently, within themselves. These systems can vary in the dip of their descent into the mantle, the degree of their coupling to the over-riding plate, as well as in their own stability over time. Attempts to correlate certain properties of subducting lithospheres with these characteristics have frustratingly failed to reveal strong patterns. In an attempt to shed a little more light on the factors that influence the contrasts between subducting systems, we use 2D computational modeling techniques utilizing viscoelastic-plastic rheologies to create and evolve simulations of subduction zones. In addition to exploring differences in the properties of the subducting plate such as plate age, crustal thicknesses, and plate velocity, we also explore the role of the properties of the over-lying plate such as its mantle wedge viscosity, thermal structure, and relative velocity. Furthermore, we introduce and examine the role of material phase changes in the contact region of the converging plates as we implement proxies for such processes as hydration and eclogitization. From our early results, we see that the material properties of the contact region between the plates can strongly influence subduction zone geometry and stability. Starting from a state of subduction initiation between hypothetical oceanic and continental plates, we observe both normally and shallowly dipping slabs. Widely varying constellations of parameters sometimes result in very similar subduction geometries. For example, old, elastically thick oceanic lithosphere may sometimes generate a very similar shallowly subducting profile as that of a younger plate subducting very slowly beneath a continent. As with modeling all poorly characterized systems in less than the full dimensionality of the system itself, many of our specific results may ultimately be less than relevant to actual observable systems. However, the consistency at which the material properties of the over-riding plate and the contact zone itself appear to affect subduction zone evolution in our models does suggest that attempting to correlate plate properties with subduction zone characteristics may require taking the nature of the over-riding plate into consideration as well as that of the down-going plate.