日本地球惑星科学連合2015年大会

講演情報

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-CG57] 変動帯の構造・進化とダイナミクス

2015年5月28日(木) 11:00 〜 12:45 国際会議室 (2F)

コンビーナ:*竹下 徹(北海道大学大学院理学院自然史科学専攻)、佐藤 比呂志(東京大学地震研究所地震予知研究センター)、尾鼻 浩一郎(海洋研究開発機構 地震津波海域観測研究開発センター)、西村 卓也(京都大学防災研究所)、深畑 幸俊(京都大学防災研究所)、加藤 愛太郎(名古屋大学大学院環境学研究科)、武藤 潤(東北大学大学院理学研究科地学専攻)、佐藤 活志(京都大学大学院理学研究科地球惑星科学専攻)、小平 秀一(海洋研究開発機構 地球内部ダイナミクス領域)、鷺谷 威(名古屋大学減災連携研究センター)、石山 達也(東京大学地震研究所)、松原 誠(防災科学技術研究所)、池田 安隆(東京大学大学院理学系研究科地球惑星科学専攻)、座長:渡辺 了(富山大学大学院理工学研究部)

11:15 〜 11:30

[SCG57-18] NIED国際地震観測網データを用いて推定したフィリピン地殻内応力の空間変化: フィリピン断層の不均質すべりによる影響

*吉田 圭佑1プリード ネルソン1福山 英一1 (1.防災科学技術研究所)

キーワード:フィリピン, メカニズム解, 応力テンソルインバージョン, 応力モデリング

A detailed knowledge of the stress state of crust is a key factor to understand earthquake occurrence. Previous studies showed that the stress patterns are primarily controlled by the plate boundary forces [e.g. Zoback et al., 1989]. However, it has also be pointed out that observed complicated stress field is affected by other factors. To address this point it is important to investigate the stress state of various tectonic regions around the world.
The Philippines is located in a region of high crustal seismicity, bounded to the east and west by oblique subduction zones. In the last few years, seismic and volcanic observation network have substantially improved in the Philippines in particular during the SATREPS cooperative project ″Enhancement of Earthquake and Volcano Monitoring and Effective Utilization of Disaster Mitigation Information in the Philippines″ between PHIVOLCS and NIED. Using waveform from regional broadband seismic stations, moment tensor solutions are routinely determined at NIED. In this study, we investigated the stress states in the Philippine archipelago by using those moment tensor solutions.
As a first step, we classified focal mechanisms into three groups: 1) events within the overriding plate, 2) those along the interplate, and 3) those within the subducting plate. Then we applied the stress tensor inversion method developed by Michael (1987) to focal mechanisms within the overriding plate. The estimated σ1-axis is oriented to WNW-ESE, which is parallel to the slip vectors of the interplate events.
In order to investigate the spatial distribution of stress across the archipelago, we performed additional stress tensor inversions by dividing the entire region in sub-regions eastern and western of a stripe containing the NNE-SSW striking Philippine fault, which is a 1200km long strike-slip fault cutting through the islands. Additionally the region was subdivided along the strike of the fault. In the central and eastern sub-regions, σ1-axis are parallel to the orientations of relative plate motions. On the other hand, in the western region, orientations of σ1-axis are significantly different from those of the relative plate motions. In particular, the orientations of σ1-axis in southern part (Bohol) are substantially different from those of the relative plate motions. Also, the orientations of σ1-axis in northern part (Mindoro) are different from those in the entire Philippines.
The σ1-axes and σ3-axes in Bohol and Mindoro are opposite. The σ1-axis in Bohol and σ3-axis in Mindoro are almost parallel to the strike of the Philippines fault, which is consistent with the stress produced by left-lateral slip on the central part of the Philippines fault. This range corresponds to the Masbate fault characterized by a larger number of moderate-sized earthquakes, while the northern and southern extensions correspond to the segments which have generated several historical earthquakes [Bessana & Ando, 2005]. Our study suggests that more strain is released along the Masbate fault by moderate-sized events, while the northern and southern extensions are accumulating more strain which may be released as large earthquakes in the future. From the calculated stresses produced by the Philippines fault, based on dislocations in an elastic half-space model [Okada, 1992], as well as assuming a interplate coupling, we successfully reproduced the spatial features of stress orientations.
We also propose that a model considering the bending deformation of the overriding plate is also able to roughly explain the observed σ1-axes, based on a 2D finite element modeling, although it does not explain the strike-slip stress regime. By taking account of oblique subduction effects, we may reproduce the observations in more detail. However, we prefer the strike-slip on the Philippines fault as the cause of stress variation, because this simple model alone can sufficiently explain all the observations.