3:30 PM - 3:45 PM
[MIS16-07] Spatial distribution of 17th century tsunami deposits and reconstruction of tsunami behavior in the Yufutsu plain, Hokkaido, Japan
Keywords:tsunami deposits, Yufutsu, magnetic fabric, grain size characteristics, spatial analysis
The inland fining and thinning tendencies of tsunami deposits formed by tsunami run-up on coastal lowland are well known. Topography often affects tsunami behavior and remarkably changes the deposit thickness and grain size distribution. Therefore, comprehending how topographic relief affects sedimentary characteristics is essential for reconstructing past tsunami behavior from subsurface deposits. In this study, we reconstructed past tsunami behavior that run-up on coastal lowlands with swamps and abandoned river channels, using spatial analyses of properties of tsunami deposits (namely, thickness, grain size characteristics, anisotropy of magnetic susceptibility, and paleo-topography). Tomakomai City has a 10 km-long tsunami flood zone for maximum inundation (Japan Cabinet Office, 2020). However, studies have reported only 1.6 km available for the maximum distribution of 17th century tsunami deposits; there is an evident difference between both the values. In this study, using X-ray CT analysis, the consistency between a tsunami simulation and the sediment distribution of tsunami deposits was evaluated.
The field survey was conducted within 2.2 km of the coastline on the left bank of the Abira River. Tsunami deposits were collected from 86 sites using a Geoslicer, peat sampler, and core boring and then described. Subsequently, the samples were transported to the laboratory. Lithological descriptions, sedimentary unit classifications, and X-ray CT, grain size, and paleo-current analyses were conducted in the laboratory. Considering the spatial distribution of the above properties of tsunami deposits and (namely, lithology, thickness, and grain size characteristics) and paleo-currents estimated using magnetic fabric, tsunami behavior was reconstructed. The 17th century tsunami in Yufutsu entered from the present river mouth area and then inundated the swamp, which is a lower area compared to sand dunes. The tsunami also inundated into an abandoned river channel that was connected from the swamp. The high-energy tsunami transported numerous sediments that formed thick tsunami layers in the swamp and abandoned channels. Subsequently, the tsunami spilled over from the swamp and abandoned channels along with rising water levels. The estimated paleo-current directions were northwest (spilled over from the swamp) and north to west (spilled over from the abandoned channels). The run-up tsunami inundated at least 2 km from the present coastline. Inland fining and thinning tendencies were formed until this time because of the decrease in tsunami energy. As the tsunami flooding area was almost flat and the coastal sand dunes were seawards, the tsunami backwash flow did not occur. Recognizing no backwash units in the tsunami deposits supports this interpretation. Consequently, the second tsunami ran up in the marsh area, and a small following tsunami reached around the swamp; this is supported by the multiple units observed in the tsunami deposits in only the swamp area.
Considering the inland distribution limit of the tsunami deposits, obtained using X-ray CT analysis, we conclude that the 17th century tsunami deposits are distributed until approximately 2 km from the present coastline. In this area, a maximum inundation of 7 km was assumed; however, the actual value was approximately 30% of the assumed value.
The field survey was conducted within 2.2 km of the coastline on the left bank of the Abira River. Tsunami deposits were collected from 86 sites using a Geoslicer, peat sampler, and core boring and then described. Subsequently, the samples were transported to the laboratory. Lithological descriptions, sedimentary unit classifications, and X-ray CT, grain size, and paleo-current analyses were conducted in the laboratory. Considering the spatial distribution of the above properties of tsunami deposits and (namely, lithology, thickness, and grain size characteristics) and paleo-currents estimated using magnetic fabric, tsunami behavior was reconstructed. The 17th century tsunami in Yufutsu entered from the present river mouth area and then inundated the swamp, which is a lower area compared to sand dunes. The tsunami also inundated into an abandoned river channel that was connected from the swamp. The high-energy tsunami transported numerous sediments that formed thick tsunami layers in the swamp and abandoned channels. Subsequently, the tsunami spilled over from the swamp and abandoned channels along with rising water levels. The estimated paleo-current directions were northwest (spilled over from the swamp) and north to west (spilled over from the abandoned channels). The run-up tsunami inundated at least 2 km from the present coastline. Inland fining and thinning tendencies were formed until this time because of the decrease in tsunami energy. As the tsunami flooding area was almost flat and the coastal sand dunes were seawards, the tsunami backwash flow did not occur. Recognizing no backwash units in the tsunami deposits supports this interpretation. Consequently, the second tsunami ran up in the marsh area, and a small following tsunami reached around the swamp; this is supported by the multiple units observed in the tsunami deposits in only the swamp area.
Considering the inland distribution limit of the tsunami deposits, obtained using X-ray CT analysis, we conclude that the 17th century tsunami deposits are distributed until approximately 2 km from the present coastline. In this area, a maximum inundation of 7 km was assumed; however, the actual value was approximately 30% of the assumed value.