JpGU-AGU Joint Meeting 2020

講演情報

[J] ポスター発表

セッション記号 S (固体地球科学) » S-SS 地震学

[S-SS15] 地震発生の物理・断層のレオロジー

コンビーナ:吉田 圭佑(東北大学理学研究科附属地震噴火予知研究観測センター)、岡崎 啓史(海洋研究開発機構)、金木 俊也(京都大学防災研究所)、野田 博之(京都大学防災研究所)

[SSS15-P34] コーダスペクトル比法に基づく2016年熊本地震の余震の震源パラメータ推定とその特徴

*前迫 直人1松本 聡2光岡 郁穂12016年熊本地震合同地震観測グループ (1.九州大学大学院理学府地球惑星科学専攻、2.九州大学大学院理学研究院附属地震火山観測研究センター)

キーワード:2016年熊本地震、S波コーダスペクトル比法、震源スケーリング、応力降下量

After the 2016 Kumamoto Earthquake occurrence, a hypocenter distribution along the strike of the Hinagu Fault was found in the aftershock activity deeper than 13 km [Mitsuoka et al., 2019, SSJ fall meeting 2019]. The alignment of the hypocenter has not seen in the distribution before the Kumamoto earthquake. In addition, the activity is located above a low resistivity zone at depth of 18 km [Aizawa et al., 2019], suggesting that the cause of the special activity relates to the resistivity structure.

In order to understand the rupture characteristics of the events, it is essential parameters such as seismic moment (M0), corner frequency (fc) and stress drop (Δσ). We estimated M0 and fc for 17 aftershocks of the 2016 Kumamoto Earthquake by using a S-wave coda spectral ratio method developed by Somei et al. [2014] and calculated stress drop based on a circular crack model. Then we discussed the spatial variation in stress drops. We investigated the parameters at two groups that were divided the earthquakes into the northern group (11 events) and the southern group (6 events). The parameters for an event were obtained from a processing that we took the spectral ratio of the observed seismogram at a station between an event pair and then stacked among 13 stations to stabilize the ratio. Using the ratio with signal-noise ratio of 2 or more, we estimated seismic moment ratio to the largest event among event of a group and corner frequencies of event pairs by nonlinear least square method.

As the result, we could see the seismic moment and the corner frequency follow the relationship of M0∝fc-3. By adopting seismic moment of the largest earthquake in a group from F-net catalog and assuming a circular clack fault model, we obtained the stress drop [Eshelby, 1957; Brune, 1970,1971]. Stress drops for all 17 events varied between 3.5 and 17 MPa. These were within the range of stress drop for inland earthquake in Japan [Somei et al., 2014]. On the other hand, comparing with similar magnitude earthquakes in Somei et al. [2014], we found stress drops in this study were larger than their results.

In the northern group, we found positive correlation between seismic moment and stress drop (correlation coefficient = 0.78). Geometric average of the stress drops in each group was 6.8 MPa in the northern group, and 12 MPa in the southern group. This is consistent with the previous study that suggest aftershocks in this area caused by shear stress of several MPa to several tens MPa [Mitsuoka et al., 2019, JpGU meeting 2019].

Acknowledgement: Hi-net and F-net were provided by National Research Institute for Earth Science and Disaster Resilience (NIED), Japan. Seismogram data were provided by Hi-net (NIED) and moment tensor solutions were provided by F-net (NIED). The unified earthquake catalog was produced by Japan Meteorological Agency (JMA).