著者

相田 勇

出版者

東京大学地震研究所

雑誌

東京大学地震研究所彙報 (ISSN:00408972)

巻号頁・発行日

vol.56, no.4, pp.p713-730, 1981

被引用文献数

8

---

南海道沖におこる巨大地震による津波の波源断層モデルを数値実験によって検討した.1946年南海地震を東西二つの断層によってあらわす.今回求められた断層モデルの特徴は,西側断層の西端が,足摺岬の東側約30kmにあること,東側の断層面の傾斜角が10°と低角であることである.また西側断層による海底変動継続時間を断定することはむずかしいが,3~10分程度と,普通の地震より長い方が,計算波形と実際記録の一致がよい.1854年安政南海地震津波の波源断層モデルは,南海地震の断層とくらべて,東側断層が北側に位置し,西側断層が30km長く求められた.また,ずれの量は約15%大きい.このモデルで大阪の津波の高さは2.4mとなるが,従来の推定値よりやや低い.しかし海岸においてこの程度の高さがあるならば,堀に津波が侵入する際,多くの船を上流に押し込み,地震をおそれて船上に避難していた人多数を犠牲にする可能性は十分あると考えられる.1707年宝永津波の場合は,足摺岬付近の津波の高さが,安政津波の約1.5倍もあり,また室戸岬,高知の地盤変動量が安政地震の際の約2倍といわれている.このため安政津波と相似形の波源断層モデルではよく近似できない.ここでは西側断層を更に二つに分割して三つの断層面をもうモデルによってこれを説明した.Source models of past tsunamis generated off the Pacific coast in the Nankaido district are examined by the trial and error method of numerical experiments on the basis of seismic fault models. The fault model for the 1946 Nankai earthquake consists of the eastern and the western fault planes. The peculiarity of this model is that the western margin of the fault is located 30 km eastward of Ashizuri-Misaki and the dip angle of the eastern fault plane is as low as 10 degrees. It may be difficult to define uniquely the duration time of the bottom deformation from the results of present numerical experiments. The prevailing speculation, though, seems to be that the duration time of the western part is 3 to 10 minutes, which is slower than that of the eastern part. The reliability factor, k, of the model, a logarithmic standard deviation of the ratios of observed and computed values for five reference stations, is 1.12. The fault model for the 1854 Ansei-Nankai tsunami is 30 km longer in the western fault and 15% larger in the slip displacement than that for the 1946 tsunami. The computed tsunami height at Osaka is 2.4 meters, which is smaller than the published estimated value, 2.5~3 meters. However, since Osaka is situated on the delta of the Yodo river and has many small canals, a tsunami 2.4 meters high at shore may invade canals as waves like bores and do tremendous damages to many boats and bridges, just as described in old documents. In the 1707 Hoei-Nankai tsunami, the tsunami heights at the southwestern region of Shikoku were about 1.5 times higher and the uplift at Muroto-saki about 2 times larger than those of the 1854 tsunami. Therefore, the fault model similar to the 1854 earthquake cannot explain the above characteristics. A model having three separate faults is proposed for this tsunami.