11:00 AM - 1:00 PM
[MIS13-P04] Estimate of a tsunami source that can hit the Amami Islands
Keywords:Ryukyu Trench, Tsunami simulation, crustal deformation, Tsunami deposit
Paleo earthquake and tsunami along the Ryukyu Trench have been studied for 50 years. In the southern part of the Ryukyu Trench, it is revealed by coral boulders that tsunamis have repeatedly occurred (e.g., Goto et al. 2011; Araoka et al. 2013). On the other hand, tsunami boulders are not distributed in the Amami and Okinawa Islands along the northern part of the Ryukyu Trench, indicating this area has not been experienced large tsunamis. In addition, because coastal lowlands are poorly distributed along the Ryukyu Trench, there are few studies on sandy tsunami deposits, especially around the Amami Islands (Goto, 2017). However, historical records show that earthquakes around the Amami Islands in 1911 and 1995 caused tsunamis, and up to 10 m high tsunami hit the coastal areas of Amami-Oshima and Kikai Islands (Tsuji 1997; Goto 2013). There are several coast terraces formed by seismic crustal deformations occurred in the interval of 1500–2000 years for the last 6000 years in Kikai Island (Goto 2017). In addition, due to seismic crustal deformation, it is considered that there is a uplifting tendency in the Kasari Peninsula of Amami-Oshima Island and a subsidence tendency in the west of Naze Bay (Ikeda, 1977). These facts indicate that earthquakes and tsunamis have repeatedly occurred around the Amami Islands. Therefore, it is important to clarify the tsunami inundation history for the last several thousand years in this area to verify whether a huge tsunami has occurred in the northern part of the Ryukyu Trench.
This study conducted a preliminary coring survey using a gauge auger in coastal lowlands located on the Pacific side of the Amami Islands, Kagoshima Prefecture. As a result of the survey, the 1 cm thick sand layer was identified at 4.5 m depth in the up to 5 m deep sediment core mainly composed of organic-rich mud at the coastal marsh in the eastern part of the Kakeroma Island, south of Amami-Oshima Island. This sand layer is bounded by sharp upper and lower contacts with the surrounding organic-rich muds, implying this layer was formed by a sudden event. It has not been examined whether this sand layer was formed by a tsunami or not since we corrected the sediment core only at one site and have not done any analysis.
In the tsunami numerical simulation, we estimated the tsunami source of a tsunami that could inundate the coastal marsh in the eastern part of Kakeroma Island, where the event sand layer was identified. Numerical simulation was carried out with 32 patterns by JAGURS (Baba et al. 2015) using a single low-angle reverse fault model. Fault ruptures were set at the northern, southern, and intermediate parts off the coast of the Amama-Oshima Island, assuming a trench-type earthquake. After the simulation, we output and evaluated the inundation area at the coastal marsh and the amount of crustal deformation at the Amami Islands. Comparing the results of the northern, southern, and intermediate models, the intermediate model produced relatively high tsunami, but the coastal marsh was not inundated well. Then, we changed each parameter of the intermediate model. The coastal marsh was widely inundated when the slip amount was 20 m or more, although Kasari Peninsula, located at the northern end of the Amami-Oshima Island, was subsided. This is inconsistent with the geologically known uplift evidence of this area. On the other hand, the models in which fault width and depth were changed, the coastal marsh was not inundated well, and no model coincided with the geologically known crustal deformation. As a result, It is highly possible that the trench-type single fault rupture did not produce a tsunami that inundated the coastal marsh of the Kakeroma Island. In the future, it is necessary to conduct additional field investigations at the coastal marsh to confirm the continuity of the event sand layer and evaluate its formative event based on analyses. Then, we reconsider the models for numerical simulations, including an intraplate high-angle reverse fault model.
This study conducted a preliminary coring survey using a gauge auger in coastal lowlands located on the Pacific side of the Amami Islands, Kagoshima Prefecture. As a result of the survey, the 1 cm thick sand layer was identified at 4.5 m depth in the up to 5 m deep sediment core mainly composed of organic-rich mud at the coastal marsh in the eastern part of the Kakeroma Island, south of Amami-Oshima Island. This sand layer is bounded by sharp upper and lower contacts with the surrounding organic-rich muds, implying this layer was formed by a sudden event. It has not been examined whether this sand layer was formed by a tsunami or not since we corrected the sediment core only at one site and have not done any analysis.
In the tsunami numerical simulation, we estimated the tsunami source of a tsunami that could inundate the coastal marsh in the eastern part of Kakeroma Island, where the event sand layer was identified. Numerical simulation was carried out with 32 patterns by JAGURS (Baba et al. 2015) using a single low-angle reverse fault model. Fault ruptures were set at the northern, southern, and intermediate parts off the coast of the Amama-Oshima Island, assuming a trench-type earthquake. After the simulation, we output and evaluated the inundation area at the coastal marsh and the amount of crustal deformation at the Amami Islands. Comparing the results of the northern, southern, and intermediate models, the intermediate model produced relatively high tsunami, but the coastal marsh was not inundated well. Then, we changed each parameter of the intermediate model. The coastal marsh was widely inundated when the slip amount was 20 m or more, although Kasari Peninsula, located at the northern end of the Amami-Oshima Island, was subsided. This is inconsistent with the geologically known uplift evidence of this area. On the other hand, the models in which fault width and depth were changed, the coastal marsh was not inundated well, and no model coincided with the geologically known crustal deformation. As a result, It is highly possible that the trench-type single fault rupture did not produce a tsunami that inundated the coastal marsh of the Kakeroma Island. In the future, it is necessary to conduct additional field investigations at the coastal marsh to confirm the continuity of the event sand layer and evaluate its formative event based on analyses. Then, we reconsider the models for numerical simulations, including an intraplate high-angle reverse fault model.