Lateral magma flow in sill-complexes: towards a paradigm shift in volcanology

Abstract:

The structure of magma plumbing systems controls the distribution of volcanism, thereby influencing continental break-up and passive margin evolution. However, delimiting the structure of entire plumbing systems is difficult because: (1) intrusion networks cannot be directly accessed at active volcanoes; (2) field outcrops are limited by exposure; and (3) the resolution of geophysical data imaging the sub-surface is restricted. As a result, models involving the vertical transfer of magma in dikes, which extend from a melt source to overlying reservoirs and eruption sites, dominate the volcanic literature. Whilst there is evidence supporting the existence of vertically stacked plumbing systems, we compile a series of field- and seismic reflection-based case studies documenting the importance of extensive lateral magma transport (up to 4100 km) within sill-complexes. Most of these sill-complexes are emplaced into sediment-filled rift basins (e.g., Rockall Basin, NE Atlantic; Ceduna Sub-basin, offshore southern Australia; Karoo Basin, South Africa). There is also evidence that some sill-complexes occur within crystalline, continental crust (e.g., in the Yilgarn Craton, Australia). The case studies presented demonstrate that sill-complex emplacement is largely controlled by host rock lithology and structure. Sill intrusion is accommodated through roof uplift or, alternatively, via non-brittle processes (e.g., porosity reduction induced by host rock fluidization) that may not deform the overburden. The full or partial accommodation of magma by space-making mechanisms other than roof uplift means that intruding magma volumes may be underestimated by or completely hidden from ground deformation analyses. We show that plumbing systems need not be vertically stacked. Magma can instead be transported within laterally extensive (up to 3 × 10⁶ km²) sill-complexes, promoting the development of volcanoes that do not overlie the melt source.