日本地球惑星科学連合2023年大会

講演情報

[E] 口頭発表

セッション記号 S (固体地球科学) » S-CG 固体地球科学複合領域・一般

[S-CG45] Science of slow-to-fast earthquakes

2023年5月24日(水) 15:30 〜 16:45 国際会議室 (IC) (幕張メッセ国際会議場)

コンビーナ:加藤 愛太郎(東京大学地震研究所)、山口 飛鳥(東京大学大気海洋研究所)、濱田 洋平(独立行政法人海洋研究開発機構 高知コア研究所)、Yihe Huang(University of Michigan Ann Arbor)、座長:伊藤 喜宏(京都大学防災研究所)、西川 友章(京都大学防災研究所)

16:00 〜 16:15

[SCG45-08] Insights into the relationship between seismicity and relief of the plate interface near the trench at the Guerrero seismic gap, Mexico

*伊藤 喜宏1大柳 修慧2Plata Martinez Raymundo3、Flores Ketzallina2Garcia Emmanuel Soliman1 (1.京都大学防災研究所、2.京都大学大学院理学研究科、3.海洋研究開発機構)

キーワード:テクトニック微動、微小地震、残差重力異常

Detecting the spatial and temporal distributions of slow and fast earthquakes offshore from subduction zones has become a fundamental task in seismology. Ocean bottom seismometers (OBSs) are useful for measuring the activity of both fast and slow earthquakes on the ocean floor away from the coastline. To detect fast and slow earthquakes in the western part of the Guerrero seismic gap of Mexico, we deployed 9 to 10 OBSs in two observation periods lasting one year each, from 2017 to 2018 and from 2018 to 2019. The detection of fast earthquakes and their P- and S-arrival picking were performed with a machine learning technique, the Earthquake Transformer (Mousavi et al., 2020). The Earthquake Transformer was applied to the continuous trace of the OBS records to pick P and S arrivals at each station. After applying phase association for all the picked phases, we calculated their hypocenters. Slow earthquakes, especially shallow tectonic tremors, were detected with a modified envelope correlation method (Mizuno and Ide, 2019; Plata-Martinez et al, 2021). We proposed a silent zone, deficient in fast and slow seismicity, located at the western part of the Guerrero seismic gap, where we also identified a negative anomaly of residual gravity and topography, which we interpret as variations in the geometry of the subducting plate surface (Plata-Martinez et al., 2021). One of the major motivations for reanalyzing the OBS data with the AI-based detector is to test the hypothesis whether the geometry of the plate interface exerts some control on the occurrence of slow-to-fast earthquakes. We compared the seismic activities of both fast and slow earthquakes with the residual gravity and typography anomalies, which could be related to deflections of the plate boundary. The results show that neither small fast nor slow earthquakes have occurred in the areas where the plate boundary could be deflected in a convex or concave shape. The convex area near the trench axis is a particular area where small tsunami earthquakes have occurred. The silent zone is confirmed to be a region of low seismic activity from the results obtained using the new catalog of microearthquakes obtained through machine learning.