The formation of silicic lava by pyroclast sintering

Abstract

Silicic volcanoes produce some of the largest eruptions on Earth which can have devastating impacts on widespread areas as well as global environmental consequences. These eruptions are also lava-forming, and can often show explosive effusive hybrid activity, as observed at the recent eruptions of Cordón Caulle (2011-12) and Chaitén (2008-09), both in Chile. Unfortunately, silicic eruptions are common in the rock record, and this means that our understanding of them needs to drastically improve if we are to properly address their potentially significant hazards. One outstanding question surrounds silicic eruption hybrid behaviour, where explosive and effusive mechanisms are coincident in space and time; how does this occur? Ash and other pyroclasts which are entrained in large eruption plumes are the products of magma fragmentation and are very different in terms of eruption dynamics and textures compared to their lava counterpart. So how can two extreme end-member behaviours occur at the same vent, simultaneously from the same source magma? This thesis investigates the formation of lava within the framework of a new model for silicic eruptions; the so-called ‘cryptic fragmentation model’. The cryptic fragmentation model suggests that lava is the by-product of magma fragmentation and involves the sintering and welding of ash and pyroclasts in the volcanic conduit. Lava that is formed in the conduit is then later extruded out of the vent. The focus of this investigation is on a 24 ka, rhyolitic eruption at Hrafntinnuhryggur, Krafla, Iceland, an exceptionally preserved ridge that has conduit deposits accessible to depths up to 70 m below the ridge surface. Further, these conduit deposits are in stratigraphic continuity with surficial lava. This thesis broadly divides into three detailed investigations:

(1) A detailed examination of the micro- and macro-texture and dissolved volatile contents in the conduit rocks located at Hrafntinnuhryggur. The principal finding is that there is unequivocal evidence that the conduit rocks were formed from the sintering of pyroclasts, inclusion of lithic material at all scales, and open-system degassing into a fragmented conduit prior to sintering and sealing.

(2) An experimental campaign to replicate the natural textures directly using the ‘vented bubble model’ for viscous sintering of obsidian particles. A principal finding here is that small particles degas and sinter efficiently producing dense obsidian, while larger particles can internally vesiculate and sinter more slowly, complicating the dynamics. The results of this investigation includes a sintering model for general use.
(3) A micro- and macro-texture examination of the deposits found in the Hrafntinnuhryggur subaerial lavas, with a focus on both primary lava textures (e.g. flow band formation from particles) and secondary overprint textures including post-emplacement vesiculation within larger clasts and brecciation-and healing cycles.

These investigations come together around the new model ‘cryptic fragmentation’ that reframes our understanding of silicic volcanism and may help to better understand our expectations of these large-scale eruptions and what hazards form from beginning to end. In this thesis, the closing elements are a discussion of the extent to which this model is generally applicable, framed through comparisons with other localities and other suites of eruption or deposit observations. The principle conclusion is that the cryptic fragmentation model can explain a substantial amount of the existing observations at silicic eruptions worldwide.