Redeposition of volcaniclastic sediments by a tsunami 4600 years ago at Kushima City,south-eastern Kyushu,Japan

The Hyuga-nada Sea, south-eastern Kyushu, Japan, is located between a strong (Nankai Trough) and a weak interplate coupling zone (Ryukyu Trench). Over the past 400 years this area has only experienced Magnitude 7·5 earthquakes or smaller and associated small-scale tsunamis. However, this short historical record most likely does not include the full range of high magnitude, low frequency giant earthquakes that might have occurred in the region. Thus, it is still unclear whether giant earthquakes and their associated tsunamis have occurred in this region. This paper reports on a prehistoric tsunami deposit discovered in a coastal lowland in south-eastern Kyushu facing the Hyuga-nada Sea. There is a reddish-brown pumiceous layer preserved in a non-marine, organic-rich mud sequence obtained from onshore sediment cores. This layer is recognized as the ca 4600 year old Kirishima-Miike tephra (that is now placed around 4500 years ago) sourced from Mount Kirishima, southern Kyushu. Another whitish pumiceous layer is evident below the Kirishima-Miike tephra in almost all of the sediment cores. A relatively high percentage of marine and brackish diatoms is recorded within this lower pumiceous layer (but not in the surrounding muds or in the overlying Kirishima-Miike tephra), indicating a marine or beach sediment source. Plant material obtained from organic-rich mud immediately below the event layer was dated to ca 4430 to 4710 cal yr bp , providing a limiting-maximum age for this marine incursion event. The presence of marine diatoms below the event layer is probably explained by pre-seismic subsidence. An absence of the resting spore of the planktonic brackish diatom Cheatoceros and the appearance of the freshwater diatom Eunotia serra immediately above the event layer probably represents a marked change to a relatively low-salinity environment. Assuming that there were no significant local geomorphological changes, such as drainage obstruction caused by formation of a new barrier spit, it is considered that co-seismic or immediate post-seismic uplift are the most likely explanations for this notable environmental change. Based on the crustal movements noted before and after the marine incursion, this event is interpreted here as an earthquake-generated tsunami. Moreover, because of these notable seismic crustal movements the tsunamigenic earthquake probably occurred immediately offshore of the study site.