After the July 2019-June 2020 small-scale magmatic activity, surface unrest of the Nakadake first crater, which is located at the center of Aso caldera, SW Japan, had been mostly calm for fourteen months, and a lake had reformed inside the crater by late-August 2021. An eruption producing ballistic clasts and a tephra fall deposit occurred within the first crater of Nakadake at 04:44 on October 14, 2021. A large number of ballistic clasts were distributed from the west-northwestern rim of the Nakadake first crater to the southern rim of the second crater, with ballistics also reaching at least 450 m south of the center of the first crater. The largest clast (70×32×31 cm) was ejected a distance of 300 m SW of the center of the first crater. Several impact craters, which were<1 m in diameter, were observed in the surface ash layer at the crater rim. The ballistic clasts were dominated by basaltic-andesite lithic fragments of lavas and pyroclastic rocks, and were not thought to derive from a newly ascending magma. The tephra fall deposit was distributed to the southeast, extending approximately 30 km from the source crater. At the crater rim, the fall deposits were composed mainly of sand-size particles with small amounts of lapilli (<17 %), and were aggregated at sizes of a few mm to 1 cm. The aggregated muddy ash (a few millimeters in diameter) adhered to plant leaves and the surface of man-made constructions in the southeastern part of Aso caldera (4-10 km), indicating that the rising plume contained large amounts of condensed water vapor. Based on the isomass map, the total discharged mass of the October 14, 2021 eruption was calculated at approximately 2500 tons. Gray to white lithic grains (40-50 %) were dominant in the tephra deposits (0.125-0.25 mm fraction), while black to brown glass shards (8-16 %) were also observed. Although a very small proportion of glass particles appeared to be fresh, most of glass shards showed varying degrees of alteration based on microscope observation and electron micro-probe analysis. These combined lines of evidence suggest that the October 14, 2021 eruption of the first crater at Nakadake was probably a purely phreatic eruption.

In active geothermal areas, subsurface high-temperature thermal waters occasionally cause phreatic (hydrothermal) eruptions without any direct input of mass and energy from magma. So, understanding subsurface hydrothermal systems is critical to improving mitigation strategies for such hazards. The Noboribetsu geothermal area in Kuttara volcano, southwestern Hokkaido, has had repeated phreatic eruptions through the Holocene. In this study, to reveal the hydrothermal system beneath this geothermal area, we investigate (1) the chemical and isotopic compositions of thermal waters and fumarolic gases and (2) the characteristics of hydrothermally altered rocks in phreatic ejecta and around thermal water discharge areas. The chemical and isotopic features of the thermal waters indicate that the hydrothermal activity in this area is attributable to a deep thermal water with a Cl concentration of approximately 12,000 mg/L and a temperature>220 °C. The hydrothermally altered pyroclastic rocks in the phreatic ejecta often include vesicles filled with smectite, chlorite, and Ca-zeolite, implying that a low-permeability clay cap consisting of these minerals exists in the subsurface and impedes the ascent of the deep thermal water. The deep thermal water ascends partly to the shallow subsurface, causing separation of the vapor phase containing CO₂ and H₂S due to boiling, and the liquid phase discharges as neutral NaCl-type waters. In addition, absorption of the separated vapor phase by groundwater, with oxidation of H₂S, leads to the formation of steam-heated acid-sulfate waters, which cause acid leaching and alunite precipitation in the shallow subsurface. The Hiyoriyama fumaroles are derived from the vapor separated from the deep thermal water at 140 °C. Phreatic (hydrothermal) eruptions in the Noboribetsu geothermal area are assumed to have occurred due to rapid formation of a vapor phase caused by a sudden pressure drop of the deep thermal water. Because such eruptions are likely to occur in this area in the future, we should perform efficient monitoring using the constructed model of the hydrothermal system.

We carried out a temporary seismic observation using 4 ocean bottom seismographs in and around the Aira caldera to obtain the precise hypocenter distribution, especially in the Wakamiko caldera. During the observation period from July to September 2020, we observed two short-time seismic swarms that occurred inside the Wakamiko caldera and revealed the precise distribution of these hypocenters and the focal mechanism solution. In contrast to our results, the JMA unified hypocenters of the two swarms are widely distributed in the western region outside the Wakamiko caldera. This difference in the hypocenter distributions insists that the seismic observation in the sea area around the Wakamiko caldera is required to reveal the precise seismic activity in the caldera.