Caldera collapse at near-ridge seamounts: an experimental investigation

Abstract

Collapse calderas are sub-circular volcanic depressions caused by subsidence of the magma reservoir roof during an eruption. Scaled physical models of caldera collapse using flat topography have been instrumental in investigating the spatial and temporal development of calderas, in particular, two distinctive sets of concentric ring faults, one reverse and one normal. More recent analog studies have investigated the effect of non-flat topography which alters the principle stress trajectories and resulting collapse structure. This work provides the basis for investigating how naturally scaled topographic loads may affect caldera collapse in relation to shallow magma reservoirs. The objective of this study is to understand how a near-ridge seamount affects caldera collapse from both a central and offset position as the seamount migrates above the magma reservoir as a result of plate motion. We utilize scaled analog models of caldera collapse in conjunction with three-dimensional (3D) laser scanning and digital particle image velocimetry (DPIV) to investigate caldera collapse dynamics at near-ridge seamounts. Experiments using a seamount cone positioned centrally above the magma reservoir result in (1) increased subsidence along the interior outward-dipping faults and (2) a preference to more symmetric collapse patterns as indicated by the subsidence profile and structure of the caldera relative to experiments with an offset cone. When the cone is offset, the collapse is asymmetric and trapdoor in nature, with the center of greatest subsidence displaced away from the region of largest topographic load. For these latter experiments, subsidence is focused where the roof is thinnest along an initial reverse fault, followed by a transition to an antithetic graben structure. The asymmetric collapse in the experiments results in a caldera with a tilted profile. Offset calderas at near-ridge seamounts are tilted towards the ridge axis, suggesting that they may have collapsed asymmetrically similar to our experiments. Furthermore, the graben structure observed in asymmetric collapse experiments is consistent with structures observed in some natural trapdoor basaltic volcanoes.