The ca. 17 cal. ka BP eruption at Nantai Volcano, NE Japan, initially produced widespread Plinian fallout deposit (Nantai-Imaichi Tephra:Nt-I) and two overlying associated scoria flow deposits, i.e., dacitic pyroclast-rich, Shizu Scoria Flow Deposit (SZ) and andesitic pyroclast-rich, Takanosu Scoria Flow Deposit (TKS). A∼2.8m thick outcrop of the Nt-I at Nikko City, 7.5km ESE of the volcano, consists of a basal phreatic fall bed (∼2cm thick) and eleven overlying fall units (units 1-11 in ascending stratigraphic order) defined by componentry, size grading, and chemical composition of the pyroclasts. The total lack of clear boundary structures between each unit suggests that the Nt-I was generated by the pyroclasts falling from continuous eruptive column. Grain size analyses of the Nt-I shows that column height rapidly increased and reached its climax soon after the eruption began, and then oscillated slightly and declined until the end of the Plinian phase. The composition of the pyroclasts shows that the Nt-I resulted from the tapping of a stratified magma chamber, in which dacitic magma capped hybrid andesitic magma. Light-colored, microlite-free, dacitic pumice (DWP) predominates from unit 1 through unit 9. In contrast, moderately vesicular andesitic scoria (AGS) is a major constituent of units 10 and 11. Microlite-rich dacitic obsidian (DOB) is present from unit 1 through unit 3, but was not observed above unit 3. Microlite-rich dacitic scoria (DBS) is present from unit 1 through unit 8, and coexists with DOB in single pyroclast. A plausible explanation for the common eruption of a small amount of microlite-rich pyroclasts along with the predominant DWP is that the microlite-rich pyroclasts represent fragments of the degassed margins of the conduit through which the dacitic magma rose. As the eruption advanced, the passageway may have widened, and the microlite-rich magma along the conduit wall was eroded and ejected along with the DWP. The density of the DWP remained constant from unit 1 through unit 8, and then increased at unit 9. The incorporation of slightly denser, dacitic pyroclast into the column is likely to have destabilized the eruption column. The destabilization caused partial collapse of the column and generated the intra-Plinian Shizu Scoria Flow Deposit, whose particle density is similar to that of unit 9. In contrast, the ejection of dense AGS combined with the incorporation of dense lithics into the eruption column is likely to have destabilized the column, and triggered total column collapse that formed the post-Plinian Takanosu Scoria Flow Deposit.

The Asama-Maekake volcano has been active for about 10,000 years. Although the many pyroclastic fall deposits, which mainly consist of pumice fall layers, are indicative of past sub-Plinian eruptions, the stratigraphy of the deposits over the entire history of the Asama-Maekake Volcano has not yet been revealed. This is because the pyroclastic fall deposits are distributed in the various directions from the crater. Moreover, the deposits predating the 12^th century have a similar occurrence and petrography, making it difficult to correlate deposits among different localities. Therefore, little information is available for the older eruptions. The pyroclastic fall deposit in the direction not leeward of the dominant wind is focused in this study. A pyroclastic fall deposit called Miyota pumice (referred to as As-My hereafter), which is characteristically distributed in the southern direction from the crater, was targeted. The distribution of As-My has barely been mapped owing to the difficulty resulting from its poor exposure. The ¹⁴C ages of the samples of black humus soil immediately beneath As-My at two localities were dated to 5720±30yBP and 5530±30yBP. These ages are almost the same as those of the pyroclastic fall deposits As-UB (Ubagahara) and As-Kn (Kuni) distributed on the northern flank. Since the stratigraphic relation among these deposits is difficult to determine, it is not clear whether As-My, As-UB, and As-Kn are products from a single eruption or different eruptive activities. At least, it can be considered that multiple sub-Plinian eruptions occurred around 6,000 years ago. In addition, black humus soil and intercalating four pumice fall deposits including As-My at a major outcrop on the SSE flank slope (Locality M72) provided information on the activity over the last 8,000 years, back to the earliest stage of the volcano. The ¹⁴C ages of samples of black humus soil taken from immediately beneath the pumice fall deposits were 3,830±30yBP, 4,710±30yBP, 5,530±30yBP (just beneath As-My), and 7,470±30yBP. This is the first time that such sequential data concerning eruptive ages on the flank slope has been obtained. Black soil contains angular lithic fragments, which are similar to the ash grains from Vulcanian eruptions (e.g., the 2004 eruption). This suggests that Vulcanian eruptions have occurred between sub-Plinian eruptions since the early stage of the Asama-Maekake volcano. Such information from outcrops on the flank slopes with various directions is required to reconstruct the detailed eruptive history.