[S-VC32] Dynamics of volcanic eruptions and their physical and chemical processes

[SVC32-03] The origin of pyroclastic obsidian: A case study of obsidian in Kemanai pyroclastic flow from the 915 eruption of Towada Volcano, Japan

*Rikuto Hirata¹, Satoshi Okumura¹ (1.Division of Earth and Planetary Materials Science, Department of Earth Science, Graduate School of Science, Tohoku University)

Keywords:pyroclastic obsidian, open system degassing, bubble, water, vesicularity, cooling rate

Pyroclastic obsidian often erupts together with pumice during explosive eruptions (e.g., Newman et al., 1988). The pyroclastic obsidians include only a small amounts of gas bubble and their water contents are higher than those of pumices (e.g., Giachetti et al., 2020). The bubble-free or less vesiculated obsidians are thought to originate from open system degassing of hydrous rhyolitic magma during the ascent (e.g., Eichelberger et al., 1986). In contrast, high water contents of the obsidian indicate that the magma is quenched at high pressure; however, the origin of pyroclastic obsidian is unclear yet. In this study, we investigated bubble microstructure and water content in pyroclastic obsidians collected from pyroclastic flow deposit of the 915 eruption of Towada Volcano. Before bubble and water content analyses, we measured bulk rock chemical compositions of pyroclastic obsidians and pumices erupted together with the obsidians and confirmed that their compositions are similar. The bubble observation was conducted using X-ray CT and FE-SEM, and the water contents were analyzed using FT-IR microspectroscopy. The water contents of the obsidians and pumices were also measured by Karl-Fischer titration. The analytical results showed that the bubbles in pyroclastic obsidians are elongated and aligned, and they seem to be formed through breakup of strongly elongated bubble under deformation. The OH and H₂O_m contents in the obsidians measured by FT-IR microspectroscopy are 1.04±0.06 and 1.05±0.15 wt.%, respectively. The water speciation was applied to geospeedmeter (Zhang et al., 1997), resulting in the cooling rate of 0.0011–0.064 °C/s. These results imply that magma with high water content (2.10±0.20 wt.%) was deformed and cooled slowly. Based on these results, we propose that pyroclastic obsidian was generated at the conduit margin in the deep conduit (~58 MPa at 1000 °C based on VolatileCalc of Newman and Lowenstern, 2002). The magma was slowly cooled at the conduit margin via the interaction with country rock. Because of the decrease in temperature, the magma did not vesiculate after cooling and thus water content remained high. The water contents in pumices erupted together with pyroclastic obsidians were lower than those in the obsidians, because the pumices were not cooled and continuously vesiculated during the ascent and thus decompression.