Physical Processes in Subaerial and Submarine Explosive Volcanism: Case Studies from the Kermadec Arc, SW Pacific

Abstract

This research explores the dynamics of subaerial and submarine explosive silicic volcanism in the Kermadec arc (SW Pacific) by looking at the microtextures of bubbles in pumice. An increasing body of evidence from oceanic arcs shows that submarine pumice-forming eruptions are surprisingly common. Such eruptions involve silica-rich magmas, despite their oceanic setting, and many are large enough to form calderas and represent a significant hazard. This study contributes towards the understanding of conditions during silicic volcanic eruptions, using comparisons and contrasts between pyroclasts from subaerial and submarine eruptions from the same tectonic setting. Five volcanoes are investigated (Raoul, Raoul SW, Macauley, Healy and Havre), all of which have erupted dacite to rhyodacite of broadly similar chemical and physical properties within the last 10,000 years. These eruptions occurred over a range of water depths providing constraints that allow such processes to be quantified by ‘eliminating’ the variables of magma chemistry and tectonic setting. In this thesis I quantify the pyroclast bubble size distributions and number densities for subaerial and submarine pumices using an adaption of the FOAMS program [Shea et al., 2010a: JVGR 190, 271-289]. The investigation of pumices from six subaerial eruptions from Raoul volcano show there is a dearth in pyroclasts with ~65-75% vesicularity with most deposits having pyroclasts with vesicularities higher than this range (mode ~82%). It is these 65-75% vesicularity clasts, however, that have the highest bubble number density (BND) values, regardless of eruption intensity, style or degree of interaction with external water. This study suggest that this 65-75% vesicularity range is pivotal in the fragmentation of magma, with higher vesicularity clasts preserving varying degrees of post-fragmentation bubble growth and coalescence, and hence decrease in BND values. The implications from this study are broad reaching as most pyroclast bubble size studies focus on modal density clasts, which I show do not preserve characteristics of the magma at fragmentation, but instead preserve some degree of post-fragmentation expansion prior to quenching in the conduit or eruption plume. The thorough assessment of pyroclast textures from the subaerial erupted Raoul pyroclasts allows for comparisons to be made with submarine erupted pyroclasts sampled via dredging. These results permit inferences to be drawn as to the influence of both eruption rate and water depth on the eruption dynamics, with the interplay between the two playing a vital role. Results of this work challenge the existing notion of simple end-member explosive or effusive regimes and define a new intermediate eruptive style (Tangaroan) that is unique to the submarine realm. This eruptive style is neither effusive nor explosive yet generates highly vesicular, widely dispersed pumice clasts through buoyant detachment of foaming magma at vent. These magma blebs undergo further vesiculation processes whilst being transported within the water column before disintegration by hydrofracturing. In contrast, higher eruption rates cause fragmentation to occur within the conduit prior to any quenching influence of the overlying water column. The higher dynamic pressure of a significant overlying water column acting on the eruption jet inhibits rapid decompression and expansion of clasts, as it would if erupted into air, and therefore affects the vesiculation processes in the resulting pyroclasts. If the eruption jet is able to breach the sea surface, the rapid decrease in pressure is translated to the pyroclast texture as an additional rapid homogenous nucleation event. This nucleation event is seen as zones of abundant small bubbles within the complex heterogeneous groundmass texture of large, thin walled, highly coalesced and sometimes highly contorted bubbles. These distinctive differences open up the possibility of being able to fingerprint subaerial versus submarine erupted pyroclasts in ancient volcaniclastic sequences.