2:15 PM - 2:30 PM
[SCG57-03] Tsunami generating factor during the 7.3 ka eruption of the Kikai caldera inferred from the stratigraphic relationship between tsunami deposits and K-Ah ash
Keywords:Volcanic glass content, Numerical tsunami simulations
The 7.3 ka eruption of the Kikai caldera is thought to have generated a huge tsunami, but the trigger of the main tsunami is still being discussed as following opinions: (1) caldera collapse, (2) inflow of pyroclastic flow into the sea, and (3) a huge earthquake occurred simultaneously with the eruption (e.g., Geshi et al., 2017; Nanayama and Maeno, 2018; Kobayashi, 2022). In this study, we aimed to clarify the principal factor of the tsunami from the stratigraphic relationship between the co-ignimbrite ash-fall deposit (K-Ah ash) and Akahoya tsunami deposit preserved in sediment cores at coastal lowlands in Oita, Tokushima, and Wakayama prefectures. In the previous studies, a tsunami factor was inferred from deposits investigated near the caldera, but it was not easy due to the complex geological record composed of volcanic ejecta at each stage, in addition to erosion and deposition caused by multiple tsunamis. In contrast, since only the main tsunami and the K-Ah ash are expected to reach the distant area, it is expected to be relatively easy to clarify the principal factor of the tsunami. In this study, we also conducted numerical tsunami simulations assuming the largest size of (1) caldera collapse, (2) pyroclastic flow, and (3) earthquake along the Nankai Trough (Mulia et al., 2017) and discussed the principal tsunami factor together with the stratigraphic relationship.
To determine the arrival timing of the K-Ah ash and main tsunami in three coastal areas, we examined the vertical variation of volcanic glass content within 3–4 phi particles collected from sediment cores. Sandy tsunami deposits are overlaid by the sandy silt layer mainly composed of the K-Ah ash and overlie a mud layer. The volcanic glass content was ~67.0% in the tsunami deposit, indicating that the K-Ah ash is considered to have fallen during the tsunami inundation at latest. In Wakayama site, the K-Ah ash layer less than 0.5 cm thick was also found below the tsunami deposit. This ensures that the K-Ah ash fell before the arrival of the tsunami. Considering that the ash, which takes longer to move than tsunami, reached these areas before or at the same time as the tsunami, it is inferred that the tsunami occurred at the end of the eruption process. Since caldera collapse occurred at the end of the eruption process (Maeno and Taniguchi, 2007), we suggest that the tsunami generated by this eruption was more likely to be mainly caused by caldera collapse than by pyroclastic inflow.
Numerical tsunami simulations were also carried out to examine the relative tsunami size and arrival time for each coastal area. In the (1) caldera collapse model, the tsunami arrived at the coastal areas between 150 and 160 minutes since the caldera collapse started, and the maximum tsunami amplitude was 2.0–4.3 m. In the (2) pyroclastic flow model, a relatively small tsunami of 0.6–1.2 m arrived the areas after 105–130 minutes since the occurrence of the pyroclastic flow. (3) huge earthquake model resulted in a relatively large tsunami ranging from 4.2 m to 10.2 m. Although it is difficult to verify the possibility of the huge earthquake theory because its rupture zone and magnitude are unknown, the wide distribution of the Akahoya tsunami deposits could not be explained by the huge earthquake alone, considering that the tsunami deposit was identified in Tachibana Bay, Nagasaki Prefecture, which is far from the trench axis (Okamura et al., 2005). On the other hand, caldera collapse model tsunami propagated through the East China Sea and resulted in 8.8 m tsunami in Tachibana Bay. These numerical results imply that the tsunami generated by the caldera collapse is essential to explain the distribution of the Akahoya tsunami deposits in distant areas. This result supports the inference based on the stratigraphic relationship of the sediment cores.
To determine the arrival timing of the K-Ah ash and main tsunami in three coastal areas, we examined the vertical variation of volcanic glass content within 3–4 phi particles collected from sediment cores. Sandy tsunami deposits are overlaid by the sandy silt layer mainly composed of the K-Ah ash and overlie a mud layer. The volcanic glass content was ~67.0% in the tsunami deposit, indicating that the K-Ah ash is considered to have fallen during the tsunami inundation at latest. In Wakayama site, the K-Ah ash layer less than 0.5 cm thick was also found below the tsunami deposit. This ensures that the K-Ah ash fell before the arrival of the tsunami. Considering that the ash, which takes longer to move than tsunami, reached these areas before or at the same time as the tsunami, it is inferred that the tsunami occurred at the end of the eruption process. Since caldera collapse occurred at the end of the eruption process (Maeno and Taniguchi, 2007), we suggest that the tsunami generated by this eruption was more likely to be mainly caused by caldera collapse than by pyroclastic inflow.
Numerical tsunami simulations were also carried out to examine the relative tsunami size and arrival time for each coastal area. In the (1) caldera collapse model, the tsunami arrived at the coastal areas between 150 and 160 minutes since the caldera collapse started, and the maximum tsunami amplitude was 2.0–4.3 m. In the (2) pyroclastic flow model, a relatively small tsunami of 0.6–1.2 m arrived the areas after 105–130 minutes since the occurrence of the pyroclastic flow. (3) huge earthquake model resulted in a relatively large tsunami ranging from 4.2 m to 10.2 m. Although it is difficult to verify the possibility of the huge earthquake theory because its rupture zone and magnitude are unknown, the wide distribution of the Akahoya tsunami deposits could not be explained by the huge earthquake alone, considering that the tsunami deposit was identified in Tachibana Bay, Nagasaki Prefecture, which is far from the trench axis (Okamura et al., 2005). On the other hand, caldera collapse model tsunami propagated through the East China Sea and resulted in 8.8 m tsunami in Tachibana Bay. These numerical results imply that the tsunami generated by the caldera collapse is essential to explain the distribution of the Akahoya tsunami deposits in distant areas. This result supports the inference based on the stratigraphic relationship of the sediment cores.