100 years of magma evolution at Sakurajima volcano, Japan
Abstract
The Sakurajima volcano formed 26 ka ago in the southern part of Aira caldera in South Kyushu, Japan. It is known for 17 Plinian eruptions, with 4 recorded in historical time. The latest event in 1914 was the largest volcanic event of the 20th century in Japan. Recent ground deformation modelling indicates that the surface around the volcano has inflated to a similar level as in 1914. Additionally, a change in magmatic conditions is implied from volcanic activity patterns, which have shifted from intermittent Plinian and effusive eruptions to frequent Vulcanian explosions over the last 100 years. This also coincides with decreasing SiO2 compositions of eruptive deposits from pre-1914 dacitic contents to the recent basaltic andesite ejecta.
We present a comprehensive geochemical study of eruptive deposits since the 1914 Plinian eruption to determine whether geophysical observations can be linked to changes in magma dynamics for the different styles of eruption. Results allow us to build a magma system model for Sakurajima volcano in which subalkaline, pyroxene-bearing and med-K magma is generated by an addition of a 5-6% melt of upper and lower crust in varying proportions to a mantle-derived primitive basaltic magma. Our trans-crustal magma model implies that a magma reservoir consists of lenses of melt stored in a crystal mush which vary slightly geochemically depending on their depths within the crust (11-18 km inferred from geothermobarometry). The fresh supply of mafic magma causes magma mixing and an overturn of the reservoir leading to eruption. The narrow variation of Pb isotope ratios along with increasing MgO content indicate more frequent injections of fresh mafic liquids to the fractionated precursory magmas since 1914. This is also supported by observations of disequilibrium processes inferred from reverse zoning, corroded cores and sieve textures of phenocrysts. Trace element modelling provides evidence for a small degree of fractionation and substantial partial melting. The ongoing research has been testing this model further by estimating depths and temperatures of magma sources for individual samples using geothermobarometry and assessing magma residence times through geochronometry. The results will allow to link particular magmatic processes with patterns observed in geophysical monitoring.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.V13A..03Z
- Keywords:
-
- 1036 Magma chamber processes;
- GEOCHEMISTRY;
- 1199 General or miscellaneous;
- GEOCHRONOLOGY;
- 7299 General or miscellaneous;
- SEISMOLOGY;
- 8439 Physics and chemistry of magma bodies;
- VOLCANOLOGY