2:15 PM - 2:30 PM
[MIS11-03] Reconstructing tsunami behavior during the 2024 Noto Peninsula earthquake: Insights from onshore tsunami deposits in Uchiura, Noto Town, Japan
Keywords:2024 Noto Peninsula earthquake, Tsunami deposit
On January 1, 2024, a Mw 7.5 earthquake struck the Noto region of Ishikawa Prefecture (Japan Meteorological Agency, 2024). The earthquake triggered widespread tsunami inundation, causing severe damage and destruction of houses, particularly in the northeastern coastal area of the peninsula (Geospatial Information Authority of Japan, 2024; The 2024 Noto Peninsula Earthquake Tectonic Landform Research Group (The Association of Japanese Geographers), 2024). However, no tide gauge stations are located in the northeastern coastal area, and tsunami waveform data has not been recorded so that the actual scale of the inundating tsunami flows has not been understood yet. Therefore, investigating onshore tsunami behavior in this region is crucial for reconstructing the tsunami size and source by numerical simulations. This study aims to reconstruct the tsunami behavior by examining tsunami traces and deposits. A field survey was conducted in Uchiura, Noto Town, Ishikawa Prefecture, one of the most severely affected by the tsunami.
Uchiura is a coastal lowland located on the northeast coast of the Noto Peninsula. The Kurikawashiri area lies on the right bank of the Kurikawashiri River, while the Nunoura area is on the left bank. This river flows through the center of the region. In this region, the tsunami inundated up to 750 m inland, with a wave height near the coast estimated to be at least 4.4 m based on observed tsunami traces. In the Kurikawashiri area, located further inland, the bent direction of plants indicated that the tsunami ran up the river, overflowed its banks, and inundated the lowlands. Notably, significant overflow was observed in two locations: near the tsunami inundation limit along the riverbank (Loc. A; 750 m from the shoreline) and at 350 m from the shoreline (Loc. B).
Relatively thick tsunami deposits were observed near Locs. A and B. Survey pits were excavated at 15 sites in Loc. A and 13 sites in Loc. B. In the Loc. A, sandy tsunami deposits composed mainly of fine to medium sand were observed, with an average thickness of 11.3 cm. Additionally, mud drapes were present at a depth of 2 cm within the tsunami deposits, and a sub-unit structure separated by the mud drape was observed only near the riverbank. This suggests that the tsunami deposits near the riverbank were formed by at least two distinct flows. In previous studies of modern tsunami deposits, sub-unit structure has been interpreted as corresponding to the number of tsunami flows, such as run-up and backwash flows (e.g., Nanayama and Shigeno, 2006; Paris et al., 2007). Given that the lowlands at the Kurikawashiri area are flat, it is unlikely that the maximum wave overflowed the riverbank and then receded back into the river. Instead, it is highly likely that the wave stagnated in the lowlands. Furthermore, since the distribution of the upper sand layer was concentrated near the riverbank, it is considered that this layer was formed by the run-up flow of a subsequent wave rather than by the backwash flow of the maximum wave. Therefore, in Loc. A, it is estimated that the maximum wave inundated the entire lowland by overflowing the riverbank. After that, a subsequent wave inundated only near the riverbank. On the other hand, most tsunami deposits in Loc. B consisted of a single sand layer. However, the sub-unit structure containing coarse sand and pebbles was observed only at two pits near the riverbank. Around these two pits, the pre-tsunami gravelly cultivated soil was eroded on the riverbank side, and the characteristics of the eroded material closely resembled those of the coarse sand layers within the tsunami deposits. This implies that the sub-unit structure at these pits was formed by onshore erosion and re-deposition. Therefore, in Loc. B, it is highly likely that only the maximum wave overflowed the riverbank and inundated the lowlands.
In this study area, since the tsunami that ran up the river overflowed the riverbank and inundated the lowlands, the tsunami deposits exhibited a complex thickness distribution and did not show a clear landward thinning trend. In contrast, in the case of the 2011 Tohoku-oki tsunami, which was larger in scale than the 2024 Noto Peninsula tsunami, a landward thinning trend was observed under similar topographic conditions to this study (Yamada et al., 2014). Therefore, compared to large-scale tsunamis, the thickness distribution by moderate-scale tsunamis is likely to be more significantly influenced by the presence or absence of rivers. This study suggests that the landward thinning trend is not always observed in river-run-up type moderate-scale tsunamis, where high velocity and wave height are maintained inland.
Uchiura is a coastal lowland located on the northeast coast of the Noto Peninsula. The Kurikawashiri area lies on the right bank of the Kurikawashiri River, while the Nunoura area is on the left bank. This river flows through the center of the region. In this region, the tsunami inundated up to 750 m inland, with a wave height near the coast estimated to be at least 4.4 m based on observed tsunami traces. In the Kurikawashiri area, located further inland, the bent direction of plants indicated that the tsunami ran up the river, overflowed its banks, and inundated the lowlands. Notably, significant overflow was observed in two locations: near the tsunami inundation limit along the riverbank (Loc. A; 750 m from the shoreline) and at 350 m from the shoreline (Loc. B).
Relatively thick tsunami deposits were observed near Locs. A and B. Survey pits were excavated at 15 sites in Loc. A and 13 sites in Loc. B. In the Loc. A, sandy tsunami deposits composed mainly of fine to medium sand were observed, with an average thickness of 11.3 cm. Additionally, mud drapes were present at a depth of 2 cm within the tsunami deposits, and a sub-unit structure separated by the mud drape was observed only near the riverbank. This suggests that the tsunami deposits near the riverbank were formed by at least two distinct flows. In previous studies of modern tsunami deposits, sub-unit structure has been interpreted as corresponding to the number of tsunami flows, such as run-up and backwash flows (e.g., Nanayama and Shigeno, 2006; Paris et al., 2007). Given that the lowlands at the Kurikawashiri area are flat, it is unlikely that the maximum wave overflowed the riverbank and then receded back into the river. Instead, it is highly likely that the wave stagnated in the lowlands. Furthermore, since the distribution of the upper sand layer was concentrated near the riverbank, it is considered that this layer was formed by the run-up flow of a subsequent wave rather than by the backwash flow of the maximum wave. Therefore, in Loc. A, it is estimated that the maximum wave inundated the entire lowland by overflowing the riverbank. After that, a subsequent wave inundated only near the riverbank. On the other hand, most tsunami deposits in Loc. B consisted of a single sand layer. However, the sub-unit structure containing coarse sand and pebbles was observed only at two pits near the riverbank. Around these two pits, the pre-tsunami gravelly cultivated soil was eroded on the riverbank side, and the characteristics of the eroded material closely resembled those of the coarse sand layers within the tsunami deposits. This implies that the sub-unit structure at these pits was formed by onshore erosion and re-deposition. Therefore, in Loc. B, it is highly likely that only the maximum wave overflowed the riverbank and inundated the lowlands.
In this study area, since the tsunami that ran up the river overflowed the riverbank and inundated the lowlands, the tsunami deposits exhibited a complex thickness distribution and did not show a clear landward thinning trend. In contrast, in the case of the 2011 Tohoku-oki tsunami, which was larger in scale than the 2024 Noto Peninsula tsunami, a landward thinning trend was observed under similar topographic conditions to this study (Yamada et al., 2014). Therefore, compared to large-scale tsunamis, the thickness distribution by moderate-scale tsunamis is likely to be more significantly influenced by the presence or absence of rivers. This study suggests that the landward thinning trend is not always observed in river-run-up type moderate-scale tsunamis, where high velocity and wave height are maintained inland.