Home > Research > Publications & Outputs > How to degas and fragment a subglacial basaltic...
View graph of relations

How to degas and fragment a subglacial basaltic volcano: the example of Katla 1918

Research output: Contribution to conference - Without ISBN/ISSN Abstract

Publication date31/03/2016
<mark>Original language</mark>English


During subaerial eruptions, volatiles are often considered the driving force of explosive volcanism. However, for subglacial eruptions, this role is complicated by the effects of changing pressure on volatile solubility, and the additional possibility of magma-water interaction. We are using the 1918 Katla (K1918) eruption to investigate the relative roles of these processes. K1918 was the last confirmed eruption of Katla, which is located in South Iceland. It was a VEI 4 basaltic eruption that took place under the Mýrdalsjökull glacier, resulting in a 14 km high ash plume that blanketed half of Iceland in ash and one of the largest floods of the Quaternary; a jökulhlaup with discharge rates of >300,000 m3s-1. Samples were collected from both the jökulhlaup and air-fall deposits. Discrete layers were observed at both locations and were sampled from; thus we have a record of sequential deposition at both localities. H2O concentrations were determined with Fourier transform infrared spectroscopy (FTIR) and scanning electron microscope (SEM) images were acquired at various scales, and then processed using ImageJ and FOAMS to allow a detailed quantitative investigation of both bubble and crystal populations. Preliminary results suggest that the jökulhlaup samples have retained more H2O, have higher bubble number densities (BNDs) and vesicularities, and lower crystallinities compared to the air-fall samples, indicating more rapid decompression and quenching. Many of the air-fall samples have larger bubbles in the centre of the clasts, implying continued vesiculation post fragmentation. There are also textures, particularly in the jökulhlaup samples, that are suggestive of clast recycling and open system degassing, involving the transportation of both gas and ash particles, including high Si material derived from the country rock. Our interpretation of the results is that the jökulhlaup samples mainly experienced magmatic fragmentation, before the magma had reached the ice interface. This was driven by a high vesicle content and rapid decompression, however, due to volumetric confinement, clast recycling, and magma shocking also played important roles. Once the clasts were expelled, they were quenched rapidly, preserving a wealth of textures that record various magmatic processes (such as particle transportation). In contrast, we believe the vesicle-poor air-fall samples mainly experienced phreatomagmatic fragmentation, but were quickly expelled through any meltwater and instead quenched slowly in the hot plume where they were able to continue degassing to atmospheric pressure. These interpretations offer a potential explanation for the exceptionally fast melting rate of the glacier and why such a powerful jökulhlaup was created.