人工知能を活用した津波早期警戒システムの構築(Using artificial intelligence to create a tsunami early warning system)

ad

2023-04-25 カーディフ大学

カーディフ大学の科学者たちは、最新の音響技術と人工知能(AI)を組み合わせて、水中の地震を迅速に分類し、津波発生の危険性を判断する早期警戒システムを開発しました。
このシステムは、既存の津波警報システムを補完・強化するとしており、海底マイクロフォンによって捉えられた音響放射を測定し、アルゴリズムを使用して津波の大きさと規模に関する信頼性の高い情報を提供します。
この研究は、津波被害を軽減するための自然災害警報システムの強化プロジェクトの一環として、世界中の国の警報センターに導入される予定です。

<関連情報>

過去の津波シナリオに対する効果的な細長い断層音源解の数値的検証
Numerical validation of an effective slender fault source solution for past tsunami scenarios

Bernabe Gomez,Usama Kadri
Physics of Fluids  Accepted::February 28 2023
DOI:https://doi.org/10.1063/5.0144360

Flowchart for the methodology utilized from the recording of the signal to the estimation of the potential wave heights at chosen locations.

To estimate tsunami hazards, it is first necessary to have reliable data relating to the rupture characteristics, such as epicenter, fault geometry, uplift speed, and duration. We made use of a mathematical model that combines analytical and machine learning technique capable of retrieving rupture characteristics from acoustic data. The model was applied with short computational times to data recorded by the comprehensive nuclear-test-Ban Treaty organization hydrophones during four tectonic events that were reported to trigger tsunami waves. The presented inverse problem model for acoustic waves with adequate tsunami propagation tools can be used as a complementary technique alongside tsunami warning systems due to the high propagating speeds of the sound in the ocean. In this paper, the validity of the solutions provided by the inverse problem model is tested by using the calculated earthquake parameters as input to the Cornell multi-grid coupled tsunami numerical model, which, in turn, output surface wave elevations (tsunami) to be compared against deep-ocean assessment and reporting of tsunamis buoy data.

ad
1702地球物理及び地球化学
ad
ad


Follow
ad
タイトルとURLをコピーしました