PETROGENESIS OF SYN-RIFT BASALTS IN THE WEST ANTARCTIC RIFT SYSTEM: INSIGHTS FROM FINITE-ELEMENT MODELS OF EXTENSION, MELTING, AND MELT TRANSPORT

The West Antarctic Rift System (WARS) is an extensional province lying between the Transantarctic Mountains of East Antarctica and Marie Byrd Land of West Antarctic. Extension included a Late Cretaceous – Oligocene broad rifting phase spanning the Ross Sea and a later Neogene – present(?) phase where extension focused along the margins of the WARS. Syn-rift magmatism occurred along the edges of the WARS in Victoria Land and Marie Byrd Land beginning in the Eocene and becoming more abundant after ~14 Ma. This syn-rift magmatism has been attributed to decompression melting of fossil plumes or metasomatized mantle, active plumes, or small-scale convection on the rift margins. These models invoke melting of a range of lithologies, including depleted and enriched (fertile) mantle lherzolite, pyroxenite, eclogite, and hydrous sources. To evaluate syn-rift melting of different lithologies in the WARS we adapted the finite-element modeling software ASPECT to simulate melt generation and transport during extension of a rheologically layered Earth. The models include the lithosphere and upper asthenosphere where deformation is governed by dislocation and diffusion creep, respectively. We use pre-rift crust and lithosphere thicknesses of 39 km and 180 km, respectively, for East Antarctica, and 45 km and 120 km for West Antarctica. Extension is simulated by applying an outflow boundary condition of 3.5 mm/yr at each side, an inflow condition on the bottom to balance the outflow, and a free surface at the top. Temperature conditions include surface and basal temperatures of 273 K and 1666 K. Of the lithologies tested (dry depleted mantle, dry pyroxenite, dry lherzolite, and dry basalt) only basalt melted, with melting limited to no more than 2% melt fraction. To produce significantly more melt, as is observed in the WARS, either a more fertile lithology must have been present (such as hydrous sources) or mantle temperatures must have been elevated. However, previous geodynamic models have shown that the WARS structural evolution requires that the upper mantle could not have been significantly hotter than the global average in Late Cretaceous. Our results thus suggest that syn-rift melting in the WARS was produced by hydrous sources in the mantle, although we recognize that plumes arriving during the Cenozoic contributed to melting locally near Ross Island and in Marie Byrd Land.