Do turbidites provide a record of earthquakes and tsunamis? Examination by inverse analysis using deep learning neural network

2023 Fiscal Year Final Research Report

• Project/Area Number: 20H01985
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Single-year Grants
• Section: 一般
• Review Section: Basic Section 17030:Human geosciences-related
• Research Institution: Kyoto University
• Principal Investigator: Naruse Hajime 京都大学, 理学研究科, 准教授 (40362438)
• Co-Investigator(Kenkyū-buntansha): 新井 和乃 高知大学, 海洋コア総合研究センター, 特任助教 (40757754)
• Project Period (FY): 2020-04-01 – 2024-03-31
• Keywords: 巨大津波 / 巨大地震 / イベント堆積物 / 機械学習 / 堆積プロセス

Outline of Final Research Achievements

This study aims to characterize the characteristics of turbidite currents induced by earthquakes and tsunamis by inverse analysis of turbidites using a deep learning neural network (DNN). First, we validated the method with 1-D and 2-D models by both flume and numerical experiments. Next, data were collected through field surveys in the Boso Peninsula and the Japan Trench, and inverse analyses were conducted. The results showed that past turbidity currents had overwhelmingly larger sediment discharge than those measured in present-day. Thus, the possibility that turbidites in geologic records are consequences resulting from large-scale events was increased, suggesting that turbidites in the Japan Trench may indicate past mega-earthquakes.

Free Research Field

堆積学

Academic Significance and Societal Importance of the Research Achievements

本研究は、悲惨な地質災害の防災・減災へ向けて、まったく新しい手段を創造した。例えば、タービダイトからその起源となった巨大津波の規模推定を行い，その解析結果を津波対策構造物の設計に役立てることも可能である．学術的な観点から見ると、本研究の手法が汎用性を持っていることが重要である。地質学は本質的には復元の科学であり、多くの場面で逆解析が必要となる．本研究がイベント堆積物の逆解析手法を確立したことにより、例えば層序断面から海水準変動を復元するシーケンス層序学など，さまざまな分野への波及効果が期待できる。