Timing of Magma Mixing Prior to the 2011 Eruption of Shinmoedake, Japan: On the Relationship Between Magma Injection, Magma Mixing, and Eruption Triggering

Various petrological evidences indicate magma mixing often preceded volcanic eruptions. Magma injection into the associated magma chambers also often occurs prior to eruptions as evidenced by inflation of a volcanic edifice. However, the relationship between magma injection, magma mixing, and eruption triggering is unclear because injection does not necessarily cause instantaneous mixing if the injected magma is sufficiently denser than the pre-existing magma and has formed stable stratified layers. To investigate the relationship, we estimated the timing of magma mixing prior to the 2011 sub-Plinian eruptions of Shinmoedake volcano, Kirishima volcanic group, Japan, on the basis of chemical zoning observed in magnetite phenocrysts and numerical diffusion modeling. We compared the timing with that of volcanic inflation/deflation processes. The eruptive products are comprised mainly of phenocryst-rich (28 vol%) gray pumice (SiO2 = 57 wt%) with minor amount of white pumice (SiO2 = 62 wt%). We recognized two magmatic end members, low-T dacitic magma and high-T mafic magma (basalt or basaltic andesite), and hybrid andesitic magma on the basis of our petrologic studies. Gray pumice is comprised mainly of the hybrid andesitic magma. White pumice is comprised mainly of the low-T dacitic magma with mixing of small volume of the hybrid andesitic magma. Most of the magnetite phenocrysts (type-A1) were crystallized in the hybrid andesitic magma. Their zoning profiles showed considerable increase in Mg and Al contents toward the rims of the phenocrysts, due to mixing with the high-T mafic magma. We calculated the time for diffusion to form these zoning profiles to be only 0.4 to 3 days. The short time scale suggests that the mixing of high-T magma triggered the sub-Plinian eruptions. This mixing process was not accompanied by a significant change in the volume of the magma chamber because no significant crustal deformation was observed several days prior to the eruptions (Japan Meteorological Agency, 2011). We propose magmatic overturn or melt accumulation within the magma chamber as a possible process for magma mixing without significant volume change of magma chamber. Some magnetite phenocrysts (type-A0), originated from the low-T dacitic magma, showed almost no increase in Mg and Al, indicating a short period of heating. The time for diffusion of type-A0 magnetite was less than 0.4 days. This time scale corresponds to the duration of magma ascent from the magma chamber to the surface. Mafic injection into the silicic magma chamber has likely begun with the beginning of the inflation (December 2009). The injection did not instantaneously cause an eruption but could have resulted in stable stratified magma layers to form a hybrid andesitic magma (mobile layer). This hybrid andesite then formed the main eruptive component of the 2011 eruptions of Shinmoedake. For details, see Tomiya et al. (2013: Bull. Volcanol. in press).