Kuju volcano, located within Beppu-Shimabara graben central Kyushu, Southwest Japan, has been active in the recent 200,000 years. The 54 ka Handa eruption, as large as VEI 5 or 6 and the largest one of the volcano, released large-scale pyroclastic flow deposits (Handa pfd; Kj-Hd) and a wide-spread tephra (Kj-D ash and Kj-P1 pumice fall deposits) that has been reported at more than 500 km from the source. The stratigraphic relationships among the deposits from the Handa eruption are important for volcanology and disaster prevention, and have been studied in various studies, but there is no consensus on the stratigraphy. In this study, we examined the stratigraphic relationships and the eruption history based on the stratigraphic and petrographic studies around Kuju volcano, as well as on Shikoku and Honshu Islands.

As the results, the stratigraphic relationships were revealed as follows.

1) The pumice fall deposit, that has been named Kj-Yu, was previously included in Kj-D ash layers, but is revealed to be a much older ejecta than Kj-D ash, along with the tephras newly named Kj-Tb1 and 2.

2) The clay-rich layer just below Kj-D was previously considered to be soil, but it contains a large number of volcanic ash particles so that it is defined as Kj-Y ash layer.

3) Three light brown fine ash layers, newly named Kj-D-U2, 4 and 6, sandwich between the blue grey sandy ash layers i.e. Kj-D-U1, 3, 5 and 7, are revealed to be the co-ignimbrite ash derived from Kj-Hd 1, 2 and 3 pfd, respectively. It suggests that the Kj-Hd1, 2 and 3 pfd are interbedded with Kj-D-U ash layers.

4) Kj-P1 overlies on Kj-D-U7 ash layer that mantled the reworked deposit of Kj-Hd3.

5) Kj-P1 is divided into lower and upper units based on the grain-size analysis, petrography, the chemical composition of glass shards and the isopach maps. Kj-S pfd was formed in the same time as the upper unit.

Based on the results, the eruption history is assumed as follows.

Pre-Handa eruption: the activity was low and the small-scale explosive eruptions that had released the pumice and volcanic fragments in loam (Kj-Y), followed by a relatively large explosive eruption that had formed Kj-AL.

Early phase: the eruption started with phreatic eruption, sub-plinian eruption that deposited the lower unit of Kj-D ash. Subsequently, the eruption changed to vulcanian eruptions that ejected Kj-D-U. This eruption continued for a long period time. During the time, three large-scale pyroclastic flow eruptions happened and has formed Kj-Hd1, 2 and 3. Their co-ignimbrite ashes generated from the Kj-Hd pfds were deposited as Kj-D-U2, 4 and 6. Lahar were generated after Kj-Hd2 and 3 deposition. This phase was terminated by the deposition of Kj-D-U7 ash.

Late phase: the plinian plumes occurred twice and deposited lower and upper lalyers of Kj-P1. The second one is the largest plinian eruption in the whole volcano history, with a large umbrella plume producing a wide-spread tephra at more than 500 km from the source and an intraplinian pyroclastic flow (Kj-S).

In the Atosanupuri volcanic complex in the Kussharo caldera, eastern Hokkaido, Japan, short-term uplift followed by subsidence around 1994 was detected using interferometric SAR (InSAR) analysis. In this study, an InSAR time series analysis from 2014 to 2022 using the ALOS-2 satellite revealed continued long-term subsidence of the entire Atosanupuri volcanic complex. The subsidence followed an exponential trend, with a relaxation time constant of several decades. However, long-term data are required to determine future displacement convergence due to frequent temporary and unsteady stagnations and uplifts. In contrast, the northwestern part of the Rishiri lava dome showed a constant subsidence rate without fluctuations. The results of the InSAR time-series analysis from 2016 to 2020 demonstrated that a horizontal sheet-like crust (sill) located 5.3 km below the surface of the Atosanupuri volcanic complex is shrinking at a rate of −1.44 million m³/year, whereas another sill at a depth of 700 m below the surface of the northwestern part of the Rishiri lava dome is shrinking at a rate of −21,000 m³/year. Although the residuals after subtracting these pressure source models indicate displacements of a few millimeters per year, these are most likely systematic errors inherent in InSAR. The InSAR time series analysis proved to be highly accurate in capturing temporal changes and spatial distribution, even when the displacement is less than 1 cm per year, and the results were not easily confounded by various errors. Therefore, data accumulation is crucial for InSAR time-series analysis.

Submarine volcanic eruptions produce a large amount of drifting pumice around the globe at a frequency of once every several to ten years. However, there is little knowledge about what kind of eruptions produce them. Document records of large amounts of pumice washed ashore in the Nansei Islands in southwest Japan are summarized in this article, along with an assessment of their source volcanoes and eruption frequency and style. A large amount of drifting pumice has washed ashore on the Nansei Islands eight times (1778 or 1779, 1905, 1914, 1915, 1924, 1934, 1986, 2021) since the 18th century, seven of which were after the 20th century, at a frequency of several times every 100 years. This frequency is not remarkably low compared to other natural hazard events. The eruptions that provided the source of these drifting pumice were the Fukutoku-Oka-no-Ba 1904‒1905 and 1914 eruptions, the Myojin-Shou 1915 eruption, Submarine Volcano NNE of Iriomotejima 1924 eruption, the Showa Iwo-jima 1934 eruption and the Fukutoku-Oka-no-Ba 1986 and 2021 eruptions. In the 8 recorded volcanic eruptions, including the uncertain ones, 6 were due to submarine volcanic eruptions in the southern part of the Izu-Bonin Arc. It took 2 to 6 months (mostly 2 to 4 months) for drifting pumice to reach the shores of the Nansei Islands from the Izu-Bonin Arc. The eruption styles that generated a large amount of drifting pumice can be divided into three cases. (1) An eruption that ejects a large amount of pumice from the seafloor to the sea surface and causes a steam-based volcanic plume to rise at the center of the eruption. They often occur from vents with seafloor depths of several hundred meters (Submarine Volcano NNE of Iriomotejima 1924 and Showa Iwo-jima 1934), but they also occur in shallower waters (Myojin-Shou 1915). (2) An eruption that occurs in shallow water (＜50 m), with repetitive Surtseyan eruption activity that forms islands composed of large amounts of pyroclastic material, while at the same time producing large amounts of drifting pumice (Fukutoku-Oka-no-Ba 1904‒1905, 1914 and 1986). (3) An eruption that occurs in shallow water (＜50 m) and produces a volcanic plume reaching the stratosphere. This style of eruption forms an island and generates a large amount of pyroclastic material, including drifting pumice (Fukutoku-Oka-no-Ba 2021).

Volcanic bomb is one of the most common eruption products around their source craters. Although paleomagnetic studies on volcanic bombs have a potential to provide high-resolution chronology of volcanic activity, particularly when compared with the known geomagnetic secular variation records, there are only a few such studies. In this contribution, we made an attempt to determine paleomagnetic directions from large (>1 m in diameter) volcanic bombs around “Tsubakuroswa craters”, located in Azuma volcano, for evaluating the potential use of volcanic bombs for paleomagnetic dating. Six oriented mini-cores were drilled from the central part of each large volcanic bomb, five in total, located on a gentle slope a few hundred meters south from the craters. All of the mini-cores were subjected to thermal demagnetization analysis, giving a well-determined characteristic remanent magnetization (ChRM) direction for each bomb as follows: site mean declination (D_m) of 350.6‒358.0º and inclination (I_m) of 48.9‒50.8º with a 95 percent confidence limit (α₉₅) smaller than 2.4º. The ChRM directions were consistent among the bombs, supporting the availability of volcanic bombs for further paleomagnetic dating research. Referring the geomagnetic secular variation record in this area, an all-site mean ChRM direction from the five bombs (Dec＝355.5º, Inc＝50.1º, α₉₅＝1.9º) most likely accounts for the derivation of the volcanic bombs by the Meiji Era (1893 CE) eruption. Historic pictures and descriptions are consistent with and support this interpretation. Previous reports suggested that the Meiji Era eruption did not eject magmatic materials and that the last magmatic eruption of this volcano was probably in 1331 CE. However, our results suggest that magmatic eruptions might have occurred here only ca. 130 years ago and may be largely affecting the current activity of this crater area. Our study suggests that volcanic bombs are potentially useful materials for paleomagnetic studies such as dating and establishing geomagnetic secular variation records.