9:45 AM - 10:00 AM
[SSS10-08] Crustal Stress Variations in Japanese Inland Areas
Keywords:Crustal stress variations, Japanese Island, Inland, focal mechanism
Since stress is the driving force of earthquakes, it is crucial as fundamental information for improving the accuracy of estimates regarding the maximum magnitude of future earthquakes and their occurrence probability within a given period. The stress field of the Japanese Islands was first presented as a stress trajectory map by Matsuda et al. (1978) based on the structures of Quaternary volcanoes and the slip sense of active faults. As more focal mechanism solutions of earthquakes have been accumulated, the stress field has been refined and compiled into stress maps (Townend and Zoback, 2006; Terakawa and Matsu’ura, 2010; Yukutake et al., 2015; Uchide et al., 2022). These studies have revealed the spatial distribution of the stress field, showing that some regions locally deviate from the regional stress field induced by relative plate motion. Temporal variations have been observed before and after large earthquakes accompanied by significant stress changes (e.g., Hardebeck and Okada, 2018), but detecting subtle stress variations during other periods remains challenging. To clarify the stress loading mechanism of active faults, it is also necessary to understand subtle temporal variations in the stress field.
In this study, we used a method to track temporal variations in the misfit angle, following Terakawa et al. (2015) and Imanishi and Noda (in preparation). The misfit angle is defined as the angle between the slip direction of an earthquake and the tangential traction predicted by a reference stress field. Earthquakes ordinarily occur on cracks that are oriented in alignment with the reference stress field, resulting in small misfit angles. In contrast, when events that disturb the reference stress field—such as large earthquakes, slow slip events, or fluid injection—occur, cracks with different orientations become more likely to slip, leading to an increase in the misfit angle. Therefore, by calculating the spatiotemporal average of the misfit angle, we can visualize variations in the stress field. In this study, we used the 0.2°-grid stress field estimated by Uchide et al. (2022) as the reference stress field. For focal mechanism solutions, we integrated data from the Japan Meteorological Agency (JMA) unified catalog, the JUNEC catalog (Ishibe et al., 2014), and the catalog by Uchide et al. (2022). Using these datasets, we estimated stress field variations over a period of nearly 40 years, from 1985 to 2022.
An overview of the results reveals that many regions exhibit periodic variations on the order of a year. Although further detailed analysis is required, these variations suggest a possible relationship with seasonal seismicity changes associated with snow melting in spring and increased precipitation in autumn (e.g., Heki, 2003; Ueda and Kato, 2019). Additionally, regions with large misfit angles (>50°) were found to exist sporadically, with spatial scales of several tens of kilometers. These regions tend to overlap with the distribution of slab-derived deep fluids (Kazahaya et al., 2015), suggesting that intermittent injections of high pore fluid pressure from deeper levels trigger failures along cracks with orientations different from those ordinarily expected. Furthermore, we occasionally observed increases in the misfit angle along fault zones. This may be attributed to fault unlocking due to slow slip, which causes disturbances in the surrounding stress field locally. Detecting fault unlocking in active faults is challenging with geodetic observations; however, seismic data with high spatiotemporal resolution may make it possible to detect such events. Moreover, focusing on the misfit angle before M6 or larger earthquakes, we observe a tendency for it to be smaller than the long-term average. This suggests that the proposed method could help identify periods when large earthquakes are more likely to occur.
Acknowledgments: This study utilized the JMA unified catalog and the JUNEC catalog (Ishibe et al., 2014).
In this study, we used a method to track temporal variations in the misfit angle, following Terakawa et al. (2015) and Imanishi and Noda (in preparation). The misfit angle is defined as the angle between the slip direction of an earthquake and the tangential traction predicted by a reference stress field. Earthquakes ordinarily occur on cracks that are oriented in alignment with the reference stress field, resulting in small misfit angles. In contrast, when events that disturb the reference stress field—such as large earthquakes, slow slip events, or fluid injection—occur, cracks with different orientations become more likely to slip, leading to an increase in the misfit angle. Therefore, by calculating the spatiotemporal average of the misfit angle, we can visualize variations in the stress field. In this study, we used the 0.2°-grid stress field estimated by Uchide et al. (2022) as the reference stress field. For focal mechanism solutions, we integrated data from the Japan Meteorological Agency (JMA) unified catalog, the JUNEC catalog (Ishibe et al., 2014), and the catalog by Uchide et al. (2022). Using these datasets, we estimated stress field variations over a period of nearly 40 years, from 1985 to 2022.
An overview of the results reveals that many regions exhibit periodic variations on the order of a year. Although further detailed analysis is required, these variations suggest a possible relationship with seasonal seismicity changes associated with snow melting in spring and increased precipitation in autumn (e.g., Heki, 2003; Ueda and Kato, 2019). Additionally, regions with large misfit angles (>50°) were found to exist sporadically, with spatial scales of several tens of kilometers. These regions tend to overlap with the distribution of slab-derived deep fluids (Kazahaya et al., 2015), suggesting that intermittent injections of high pore fluid pressure from deeper levels trigger failures along cracks with orientations different from those ordinarily expected. Furthermore, we occasionally observed increases in the misfit angle along fault zones. This may be attributed to fault unlocking due to slow slip, which causes disturbances in the surrounding stress field locally. Detecting fault unlocking in active faults is challenging with geodetic observations; however, seismic data with high spatiotemporal resolution may make it possible to detect such events. Moreover, focusing on the misfit angle before M6 or larger earthquakes, we observe a tendency for it to be smaller than the long-term average. This suggests that the proposed method could help identify periods when large earthquakes are more likely to occur.
Acknowledgments: This study utilized the JMA unified catalog and the JUNEC catalog (Ishibe et al., 2014).