日本地球惑星科学連合2018年大会

講演情報

[EE] 口頭発表

セッション記号 A (大気水圏科学) » A-AS 大気科学・気象学・大気環境

[A-AS01] 高性能スーパーコンピュータを用いた最新の大気科学

2018年5月20日(日) 13:45 〜 15:15 302 (幕張メッセ国際会議場 3F)

コンビーナ:瀬古 弘(気象研究所)、小玉 知央(独立行政法人海洋研究開発機構)、滝川 雅之(独立行政法人海洋研究開発機構、共同)、三好 建正(理化学研究所計算科学研究機構)、座長:瀬古 弘(気象研究所)、滝川 雅之(国立研究開発法人 海洋研究開発機構)

14:45 〜 15:00

[AAS01-05] 水文土石流モデル(Hydro-debris 2D)を用いた土石流予測

★招待講演

*山敷 庸亮1 (1.京都大学大学院総合生存学館)

キーワード:土石流、NHM、水文土石流モデル二次元版

Hydro-debris2D model has been developed and improved for predicting occurrence of debris flow throughout hydrological regime changes. The model contains three components:(1) Shallow-water based surface flow modules, in order to calculate mountain zone torrential flow regimes, (2) rapid subsurface/interflow in weathered rock, and (3) debris flow components. The model has been applied into Izu Oshima Island’s debris flow event in 2013, to Hiroshima’s debris flow disaster in 2014, and to Kyushu region heavy rainfall sediment disaster in 2017. As rainfall was input, we made a comprehensive comparison between observed rainfall station datasets from AMeDAS and High-resolution NHM calculation results for first two cases. In the case of Izu Oshima, heavy rainfall and extensive surface flow occurred in the western part of the island, together with extreme interflow which may have caused the start of debris flow in the mountain wall. In Hiroshima’s case, observed rainfall reproduced occurrences of debris flow with better agreement of the disaster due to the slight changes in heavy-rainfall zone. In Kyushu zone, using only coarse rainfall data may not be sufficient to reproduce the actual sediment disaster pattern. Reanalyses of rainfall dataset is the essential to increase accuracy of their hindcast analyses. Prediction using ensemble rainfall data may be needed in order to increase the accuracy of the occurrence.