5:15 PM - 7:15 PM
[SSS04-P07] The image of ocean volcanic activities revealed by the simultaneous epicenter and depth estimation method in ocean seismicity
Keywords:Ocean seismic activities, Submarine volcanoes, Waveform data analysis, Ocean Bottom Seismometers
Compared to onshore earthquakes, the number of seismometers deployed offshore is small and sparse, and there are few permanent stations. It is believed that many offshore volcanic events have been undetected (Tepp and Dziak, 2021). For example, a series of sudden tsunami activity that occurred near Torishima in 2023 (Sandanbata et al., 2023) is an example of undetected offshore volcanic activities. Besides, the observed waveforms traversing the oceanic lithosphere tend to suffer from strong scattering and are generally difficult to measure the arrival times accurately. Therefore, it is important to develop new methods for monitoring submarine volcanic activity in detail.
In this study, we analyzed teleseismic waveform data recorded by ocean bottom seismometers, ocean bottom cables, F-net, and JMA (Japan Meteorological Agency) stations for the submarine earthquake sequence (mainshock Mw = 6.4, normal fault earthquake) that lasted for more than one week in October 2006 near Getsuyo Seamount(29.30°N, 140.46°E), located in the Izu-Ogasawara Arc. We then processed the waveform data of P and S waves and estimated the epicenter accurately by using the following three methods. 1) By converting the seismic waveform into the temporal evolution of kurtosis, we improved the accuracy of the onset time picking. 2) By using the relative P wave arrival time/S-P time between the mainshock and aftershock as the observed quantity, we estimated the relative epicenter and relative depth accurately. 3) We ensured accuracy by visualizing the consistency among the dataset and removing inconsistent data. We applied the above three methods to the data recorded from 16 stations and simultaneously determined the epicenter and the depth.
The simultaneous estimation is carried out by using least squares method. We analyzed the P wave arrival time which is sensitive to the relative epicenter. We also analyzed the S-P time which is sensitive both to the relative depth and relative epicenter. The obtained epicenter distribution was concentrated in the rift zone in the southwest of the submarine volcano, suggesting that a series of submarine earthquakes occurred due to submarine volcanic activity (Figure 1). It was also found that the seismicity concentration became noticeable particularly after 7.5 hours since the main shock. In contrast, the relative depths scatter after 7.5 hours more than before 7.5 hours.
Based on these results, we propose the following interpretation. 1) First, the normal faulting occurred with the main shock, and seismic activity was activated in the surrounding area of the fault. 2) The extensional stress generated by the normal fault caused a dike intrusion under the rift zone. Then the stress concentration occurred at the tip of the dike, where majority of the aftershocks occurred.
In this study, we analyzed teleseismic waveform data recorded by ocean bottom seismometers, ocean bottom cables, F-net, and JMA (Japan Meteorological Agency) stations for the submarine earthquake sequence (mainshock Mw = 6.4, normal fault earthquake) that lasted for more than one week in October 2006 near Getsuyo Seamount(29.30°N, 140.46°E), located in the Izu-Ogasawara Arc. We then processed the waveform data of P and S waves and estimated the epicenter accurately by using the following three methods. 1) By converting the seismic waveform into the temporal evolution of kurtosis, we improved the accuracy of the onset time picking. 2) By using the relative P wave arrival time/S-P time between the mainshock and aftershock as the observed quantity, we estimated the relative epicenter and relative depth accurately. 3) We ensured accuracy by visualizing the consistency among the dataset and removing inconsistent data. We applied the above three methods to the data recorded from 16 stations and simultaneously determined the epicenter and the depth.
The simultaneous estimation is carried out by using least squares method. We analyzed the P wave arrival time which is sensitive to the relative epicenter. We also analyzed the S-P time which is sensitive both to the relative depth and relative epicenter. The obtained epicenter distribution was concentrated in the rift zone in the southwest of the submarine volcano, suggesting that a series of submarine earthquakes occurred due to submarine volcanic activity (Figure 1). It was also found that the seismicity concentration became noticeable particularly after 7.5 hours since the main shock. In contrast, the relative depths scatter after 7.5 hours more than before 7.5 hours.
Based on these results, we propose the following interpretation. 1) First, the normal faulting occurred with the main shock, and seismic activity was activated in the surrounding area of the fault. 2) The extensional stress generated by the normal fault caused a dike intrusion under the rift zone. Then the stress concentration occurred at the tip of the dike, where majority of the aftershocks occurred.