3:00 PM - 3:15 PM
[STT42-06] Estimation of Volcanic Earthquake Source Direction in Tonga DAS Observations
Keywords:Distributed Acoustic Sensing, Volcanic Earthquake, Hunga Tonga-Hunga Haʻapai Volcano
The massive eruption of the Hunga Tonga–Hunga Haʻapai volcano (HTHH) in January 2022 caused significant damage due to resulting tsunamis and ashfall. Compared to terrestrial volcanoes, submarine volcanoes are poorly monitored because there are few observation stations. Ocean-bottom seismometers are potential tools for monitoring submarine volcanic activity. However, they are expensive, and only a few regions are equipped with real-time cable-based ocean-bottom seismometers. In this context, Distributed Acoustic Sensing (DAS) technology, utilizing existing submarine fiber-optic cables, is expected to solve the problem.
In traditional seismic array observations, two-dimensional array configurations, such as a cross and a T-shape, are employed to constrain the direction from which seismic waves arrive. In DAS observations, if the cable has bends, it can form a two-dimensional array. However, submarine communication cables are typically laid in straight lines to connect landing stations via the shortest path possible. While the use of the existing communication cable for DAS offers an advantage, their linear geometry may not be suitable for seismic observation. This study tried estimating seismic wave arrival directions using DAS data obtained with a short linear cable.
The DAS data analyzed in this study were collected from February 6 to 13, 2023, using the Tonga domestic communication cable, which was cut at 31 km from the landing station in Nukuʻalofa (approximately 40 km south-southeast of HTHH) due to damage caused by the eruption (Nakano et al., 2024, EPS). The first 20 km of the cable from the landing station is laid on a shallow coral reef, where high noise levels are observed due to ocean waves and other environmental factors. Then, the seafloor depth increases sharply, where the cable seems to be suspended, resulting in persistently high noise levels in certain sections. Consequently, the segment with sufficiently low noise levels for detecting weak seismic events is limited to approximately 3 km near the cable’s offshore end, where the cable is laid nearly straight.
Nakao et al. (2024, JpGU) focused on the low-frequency (2–4 Hz) and long-duration events, which are characteristic of volcanic tremors, and found approximately 700 events in seven days, including those with unclear P-S phases. They analyzed the direction of arrivals for a few relatively large ones, using cross-correlation methods and referring to data from on-land seismometers. However, most of the small events had been left unresolved. Here, we investigated the direction of arrivals of these small events by comparing semblance values before and after the trigger while assuming a range of slowness values.
Our results indicate that over 30% of the 700 detected events likely originated from the direction of HTHH and the surrounding volcanic chain. The detected event rate—several tens per day—is comparable to the observation at highly active on-land volcanoes. Given that these events were recorded 40 km away from volcanoes, we may say that the event rate is extremely high. Among active volcanoes in the same direction, we consider HTHH to be the source. Our findings suggest that the volcano remained highly active even a year after its massive eruption. This study shows that it is possible to detect volcanic tremors and estimate the direction of arrival using short linear DAS arrays. This result suggests the possibility of monitoring submarine volcanoes using existing submarine cables.
Acknowledgments
This study used the data obtained by the collaboration with Tonga Cable Ltd, JICA, Dr. Takao Ohminato and Mr. Takeo Yagi (ERI, U-Tokyo), Dr. Shigeaki Ono (JAMSTEC), and Tonga Geological Services, including Dr. Daisuke Suetsugu and Mr. V. Tovi. This research was supported by Science and Technology Research Partnership for Sustainable Development (SATREPS: No. JPMJSA2309), Japan Science and Technology Agency (JST)/Japan International Cooperation Agency (JICA).
In traditional seismic array observations, two-dimensional array configurations, such as a cross and a T-shape, are employed to constrain the direction from which seismic waves arrive. In DAS observations, if the cable has bends, it can form a two-dimensional array. However, submarine communication cables are typically laid in straight lines to connect landing stations via the shortest path possible. While the use of the existing communication cable for DAS offers an advantage, their linear geometry may not be suitable for seismic observation. This study tried estimating seismic wave arrival directions using DAS data obtained with a short linear cable.
The DAS data analyzed in this study were collected from February 6 to 13, 2023, using the Tonga domestic communication cable, which was cut at 31 km from the landing station in Nukuʻalofa (approximately 40 km south-southeast of HTHH) due to damage caused by the eruption (Nakano et al., 2024, EPS). The first 20 km of the cable from the landing station is laid on a shallow coral reef, where high noise levels are observed due to ocean waves and other environmental factors. Then, the seafloor depth increases sharply, where the cable seems to be suspended, resulting in persistently high noise levels in certain sections. Consequently, the segment with sufficiently low noise levels for detecting weak seismic events is limited to approximately 3 km near the cable’s offshore end, where the cable is laid nearly straight.
Nakao et al. (2024, JpGU) focused on the low-frequency (2–4 Hz) and long-duration events, which are characteristic of volcanic tremors, and found approximately 700 events in seven days, including those with unclear P-S phases. They analyzed the direction of arrivals for a few relatively large ones, using cross-correlation methods and referring to data from on-land seismometers. However, most of the small events had been left unresolved. Here, we investigated the direction of arrivals of these small events by comparing semblance values before and after the trigger while assuming a range of slowness values.
Our results indicate that over 30% of the 700 detected events likely originated from the direction of HTHH and the surrounding volcanic chain. The detected event rate—several tens per day—is comparable to the observation at highly active on-land volcanoes. Given that these events were recorded 40 km away from volcanoes, we may say that the event rate is extremely high. Among active volcanoes in the same direction, we consider HTHH to be the source. Our findings suggest that the volcano remained highly active even a year after its massive eruption. This study shows that it is possible to detect volcanic tremors and estimate the direction of arrival using short linear DAS arrays. This result suggests the possibility of monitoring submarine volcanoes using existing submarine cables.
Acknowledgments
This study used the data obtained by the collaboration with Tonga Cable Ltd, JICA, Dr. Takao Ohminato and Mr. Takeo Yagi (ERI, U-Tokyo), Dr. Shigeaki Ono (JAMSTEC), and Tonga Geological Services, including Dr. Daisuke Suetsugu and Mr. V. Tovi. This research was supported by Science and Technology Research Partnership for Sustainable Development (SATREPS: No. JPMJSA2309), Japan Science and Technology Agency (JST)/Japan International Cooperation Agency (JICA).