5:15 PM - 7:15 PM
[SSS04-P06] Seismic activity by long-term seafloor observation in the Hyuga-nada and spatiotemporal relation of fast and slow earthquakes
The Nankai Trough experiences catastrophic earthquakes as the Philippine Sea Plate subducts northwestward beneath the Eurasian Plate. Although Hyuga-nada is located at the southwestern side of the Nankai Trough, there is no historical record of earthquakes with a magnitude of 8 or greater. To better understand the seismic activity in this region, long-term monitoring of earthquakes is essential.
This study aims to enhance the understanding of seismicity in the Hyuga-nada region through precise fast earthquake localization and comprehensive analysis with slow earthquakes. The Hyuga-nada region frequently experiences slow earthquakes, including low-frequency tremors (LFTs) and very low-frequency earthquakes (VLFEs). These events are believed to occur at the plate boundary, and their relationship with fast earthquakes remains an open question. Understanding the connections between slow and fast earthquakes may provide insights into stress transfer mechanisms and contribute to a better assessment of huge earthquake in the Hyuga-nada region. Accurate hypocenter determination is fundamental to achieving this objective.
Marine observations improve accuracy for earthquake location occurring below the seafloor. Five marine seismic arrays in 2015-2016,2017-2018, 2022, 2023 and 2024 composed of pop-up-type Long-Term Ocean Bottom Seismometers (LT-OBS) were constructed in the Hyuga-nada segment for consistent seismic monitoring. Periods of the observations are approximately a half year or one year. Waveforms were extracted from collected continuous time series data and were analyzed manually to determine P-wave and S-wave arrival times, maximum amplitude, and first-arrival polarity. Additionally, arrival times of SP-waves which were converted at a velocity discontinuitybetween sediment layers and the igneous crust were identified to correct travel times considering structural heterogeneity.
Initial event locations were obtained using the method of Hirata and Matsuura (1987) with a 1-D velocity structure based onNakanishi et al. (2018). Application of station corrections improved accuracy. Subsequent relocations were conducted using the Double-Difference seismic tomography method (Zhang & Thurber, 2003, 2006) with both absolute and relative arrival times, incorporating a 3-D velocity structure model from Nakanishi et al. (2018). Next, the method of Skoumal, Hardebeck & Shearer (2024) was used to estimate fault solutions by polarities of P-wave and S/P amplitude ratio. Finally, the Frohlich method (1992, 2001) was applied to classify focalmechanisms based on the dip angles of the pressure axis, tension axis, and null axis.
The relocated events were deeper than the plate boundary and were uniformly distributed in space and time whitin the array except for swarm events. The JMA catalog implied seismic swarms as numerous earthquakes occurring within roughly one week or a day in 2017 December, 2018 January, and 2022 September. Precise location results of this study revealed that seismic events were concentrated in a narrow-localized region during each swarm period. However. the locations of these swarms varied across different periods. Tonegawa et al. (2020) determined precise positions of VLFEs using OBS data. The comparison of the spatiotemporal distribution of 2017 and 2018 swarm events and the VLFEs shows that their active areas are identical. Furthermore, the VLFEs occurred hours before the swarm events. Comparison with the tremor active area based on previous studies shows that the fast earthquake and tremor active areas also seem to overlap.
The focal solutions revealed that strike-slip and/or normal fault types were dominant except the swarm activities. The focal mechanism of swarm events shows a significantly higher proportion of thrust fault type compared to other regular earthquakes.
This study aims to enhance the understanding of seismicity in the Hyuga-nada region through precise fast earthquake localization and comprehensive analysis with slow earthquakes. The Hyuga-nada region frequently experiences slow earthquakes, including low-frequency tremors (LFTs) and very low-frequency earthquakes (VLFEs). These events are believed to occur at the plate boundary, and their relationship with fast earthquakes remains an open question. Understanding the connections between slow and fast earthquakes may provide insights into stress transfer mechanisms and contribute to a better assessment of huge earthquake in the Hyuga-nada region. Accurate hypocenter determination is fundamental to achieving this objective.
Marine observations improve accuracy for earthquake location occurring below the seafloor. Five marine seismic arrays in 2015-2016,2017-2018, 2022, 2023 and 2024 composed of pop-up-type Long-Term Ocean Bottom Seismometers (LT-OBS) were constructed in the Hyuga-nada segment for consistent seismic monitoring. Periods of the observations are approximately a half year or one year. Waveforms were extracted from collected continuous time series data and were analyzed manually to determine P-wave and S-wave arrival times, maximum amplitude, and first-arrival polarity. Additionally, arrival times of SP-waves which were converted at a velocity discontinuitybetween sediment layers and the igneous crust were identified to correct travel times considering structural heterogeneity.
Initial event locations were obtained using the method of Hirata and Matsuura (1987) with a 1-D velocity structure based onNakanishi et al. (2018). Application of station corrections improved accuracy. Subsequent relocations were conducted using the Double-Difference seismic tomography method (Zhang & Thurber, 2003, 2006) with both absolute and relative arrival times, incorporating a 3-D velocity structure model from Nakanishi et al. (2018). Next, the method of Skoumal, Hardebeck & Shearer (2024) was used to estimate fault solutions by polarities of P-wave and S/P amplitude ratio. Finally, the Frohlich method (1992, 2001) was applied to classify focalmechanisms based on the dip angles of the pressure axis, tension axis, and null axis.
The relocated events were deeper than the plate boundary and were uniformly distributed in space and time whitin the array except for swarm events. The JMA catalog implied seismic swarms as numerous earthquakes occurring within roughly one week or a day in 2017 December, 2018 January, and 2022 September. Precise location results of this study revealed that seismic events were concentrated in a narrow-localized region during each swarm period. However. the locations of these swarms varied across different periods. Tonegawa et al. (2020) determined precise positions of VLFEs using OBS data. The comparison of the spatiotemporal distribution of 2017 and 2018 swarm events and the VLFEs shows that their active areas are identical. Furthermore, the VLFEs occurred hours before the swarm events. Comparison with the tremor active area based on previous studies shows that the fast earthquake and tremor active areas also seem to overlap.
The focal solutions revealed that strike-slip and/or normal fault types were dominant except the swarm activities. The focal mechanism of swarm events shows a significantly higher proportion of thrust fault type compared to other regular earthquakes.