Abstract
The 2012 silicic eruption at Havre volcano, in the Kermadec Arc, was the largest deep subaqueous eruption observed in the last century. A data set of unprecedented richness was collected during a dedicated research cruise in 2015, including detailed bathymetric mapping and systematic in situ sampling of seafloor clastic and effusive products. This thesis characterizes the seafloor Ash and Lapilli (AL) Unit produced during the 2012 Havre eruption, with the aim of determining the effect of the water column on ascent, fragmentation, and dispersal of ash during a deep silicic subaqueous eruption. To this end, sample grain-size distributions, sample componentry, ash shape and microtextural data, major-element chemistry, and groundmass volatile contents were acquired. Results and interpretations from the AL Unit support inferences on eruption processes.
It is demonstrated that the AL Unit is a composite deposit composed of four subunits; from base to top these are Subunit 1 (S1), 2 (S2), 3 (S3), 4a and 4b (S4). Each of the subunits in the AL Unit shows distinctive grain size or componentry characteristics, different mapped dispersal limits, and specific stratigraphic relationship with the other seafloor products of the 2012 Havre eruption. Using results of subunit depositional characteristics and particle microtextural features mechanisms are inferred to explain the generation of each subunits of the AL Unit.
Subunit 1 directly overlies the Giant Pumice Unit, draping the entire study area and fining towards the NW. This deposit is composed of an average 125 to 800 µm glassy vesicular ash showing dominantly curvi-planar morphologies, in addition to lesser amounts of angular and fluidal particles. Subunit 1 is therefore inferred to have been deposited by fallout following dispersal in an eruptive plume. The plume was driven by an eruption defined by energetic fragmentation with a large component of magma water interaction, however also apparently showing a range of other fragmentation processes.
Subunit 2 overlying S1 across a gradational contact shows a deposit boundary along the northern caldera wall. To the south of the boundary S2 is heavily thickened in the caldera showing a consistent grain size. Subunit 2 is composed of 16 to 32 µm glassy vesicular ash showing dominantly curvi-planar morphologies, in addition to lesser amounts of angular and fluidal particles. This subunit is inferred to have been deposited from dilute density currents that ponded in the Havre caldera. The similarity in microtextural features to S1 and their gradational contact suggest these two subunits were generated from the same event. With density currents potentially generated off a larger eruption column. The microtextural similarity of these deposits to the GP Unit and ALB Unit suggests their eruption from the dome OP vent, while the presence of fluidal particles and energetic fragmentation indicates and explosive eruption.
Subunit 3 drapes topography in a NE-SW trend across the caldera thinning and fining towards a lava flow on the southern caldera rim. The morphological and microtextural similarity of the ash the S3 is composed of to the pumaceous carapace of the said lava suggests this was its source. By modelling the thermal plume required to generated S3 however it is shown that weakly pyroclastic activity is required to produce the wide dispersal, likely occurring synchronously with lava effusion.
Subunit 4 is composed of microcrystalline ash, the low vesicularity and high crystallinity of which suggests fragmentation from the lava flows. Subunit 4a dispersed in a NE-SW trend across the caldera is inferred to have been generated during a caldera wall collapse near the source vents of 3 lavas produced during the 2012 eruption. Subunit 4b dispersed around Dome OP in inferred to have been generated by quenching and brecciation of the lava as it was extruded.
The results presented in this thesis show that the 2012 deep subaqueous eruption of Havre volcano was a complex event, with both explosive and effusive activity occurring over several phases. The eruptive processes were significantly influenced by the water column, which affected magma rheology, magma fragmentation to produce fine ash, initial particle dispersal, and final deposition.