Magnetite Composition as a Tool for Judging Phreatic/Phreatomagmatic Eruptions: Case Studies on Usu and Miyakejima Volcano, Japan.

In 2000 A.D., two major phreatomagmatic eruptions occurred in Japan, one was at Usu Volcano and the other was at Miyakejima Volcano. For both eruptions, identification of essential materials was controversial, however, detailed observation of microtextures of eruptives and analyses of magnetite compositions enabled us to determine the essential materials (Tomiya et al., 2000; Miyagi et al., 2000). Here, we introduce the two cases. [case 1: The 2000 eruption of Usu volcano] The first and largest explosion occurred at 31 March, where ash and small amount of pumice were ejected. The pumice (Us-2000pm; Tomiya et al., 2000) was once thought to be recycled Us-1977 pumice (ejected at the 1977 eruption) because of similar appearance, bulk composition, glass composition and plagioclase composition. However, we concluded that Us-2000pm is essential, because the magnetite composition of Us-2000pm was unique among any known Usu pumices, including Us-1977. The same technique also revealed that Us-2000g (fresh glass with many micro-bubbles and microlites, account for about a half of the ash) was the essential material of the phreatomagmatic eruption. [case 2: The 2000 eruption of Miyakejima volcano] There were several major phreatomagmatic explosions at the summit since 14 July. The largest one occurred at 18 August, where large amount of ash and cauliflower-shaped bomb were ejected. The cauliflower-shaped bomb is undoubtedly essential material because it was hot at its landing. On the other hand, presence or absence of essential materials in the ashes was controversial. Again, our analysis on magnetite in the ash particles specified the essential material (Myk2000g-2; Miyagi et al., 2000), which is ``fresh glass with many microbubbles and microlites'' and accounts for a large part (10 ~ 40 %) of the ash. Textual and compositional features of Myk2000g-2 were mostly the same among the vigorous eruptions (14-15 July; 10, 13, 18 August), except that magnetite showed drastic change between 10 August and 13 August. They were initially euhedral and homogeneous (in 14-15 July and 10 August), but they became dissolved in texture and less abundant since 13 August. The dissolution of magnetite is considered to represent the change of magmatic conditions (e.g., temperature, fO₂, melt composition), and to be a precursor for the climactic explosion at 18 August and the start of unprecedented amount (tens of thousand tons per day) of SO₂ degassing at Miyakejima volcano.