Osamu SANDANBATA1,2 & Tatsuhiko SAITO 2
1.Earthquake Research Institute, The University of Tokyo
2.National Research Institute for Earth Science and Disaster Resilience
Sandanbata, O., & Saito, T. (2024). Segmented Trapdoor Fault in Kita-Ioto Caldera, Japan: Insights From Millimeter Tsunami Waves Captured by an Array Network of Ocean Bottom Pressure Gauges.
Journal of Geophysical Research. Solid Earth, 129(12), e2024JB029755. https://doi.org/10.1029/2024JB029755
Key Points
- Kita-Ioto Caldera, a submarine volcano in Japan, has recently caused Mw 5.2–5.3 non-double-couple earthquakes every 2–5 years.
- Stacking ocean bottom pressure data at ~1,000 km from the source extracts 1–2 mm tsunami signals, enabling us to reveal the trapdoor faulting mechanism of the 2017 and 2019 earthquakes.
- The detected tsunami waveforms are distinctly different between the two earthquakes, suggesting the segmentation of the intra-caldera fault system, which may result in the frequent recurrence of trapdoor faulting.
Background
In the submarine volcano of Kita-Ioto Caldera in the Ogasawara Islands, moderate earthquakes with magnitudes (M) of 5.2–5.3 have been recurring every 2–5 years (Figure 1). Our previous study (Sandanbata & Saito, 2024a) analyzed data from tsunami sensors installed in the deep-sea region of the Philippine Sea. It revealed that the earthquakes in 2008 (M5.3) and 2015 (M5.2) were caused by the rapid rupture of intra-caldera fault structures, a phenomenon known as “trapdoor faulting” (see Sandanbata et al. (2022) for the tsunami generation mechanism by trapdoor faulting). These events triggered abrupt caldera uplift and then generated tsunamis. However, for the earthquakes in 2017 (M5.2) and 2019 (M5.3), data from the tsunami sensor in the Philippine Sea were lost, leaving the causes and characteristics of these events insufficiently understood.

Detection of Micro-Tsunamis with Maximum Amplitudes of 1–2 mm
Understanding the mechanisms and characteristics of the 2017 and 2019 earthquakes is crucial for assessing the recent volcanic activity of Kita-Ioto Caldera and its long-term trends. In this study, we utilized data from the Dense Oceanfloor Network for Earthquakes and Tsunamis (DONET), a high-density tsunami observation network installed more than 900 km away from the submarine calderas (Figure 1c), to detect tsunami signals generated by these earthquakes.
Although identifying tsunami signals from individual records proved challenging (Figure 2a), we applied a waveform stacking method. This technique leverages waveform similarities across multiple records to suppress noise and amplify signals, enabling the successful detection of micro-tsunami signals with maximum amplitudes of 1–2 mm (Figure 2b). Based on their arrival times and frequency characteristics, these micro-tsunamis are attributed to the earthquakes at Kita-Ioto Caldera and were transmitted to the DONET observation area.

Characteristics of Trapdoor Faulting Repeating in Kita-Ioto Caldera
Micro-tsunamis with amplitudes of 1–2 mm pose no direct threat to coastal areas or vessels; however, their records serve as valuable data for investigating the submarine phenomena that generate tsunamis. By comparing the detected micro-tsunami records with numerical simulations of tsunami generation and propagation, we analyzed the mechanisms and source characteristics of the 2017 and 2019 earthquakes. Our findings revealed the following:
- Common mechanism for both earthquakes—trapdoor Faulting: The micro-tsunami records from both events closely matched numerical simulations that assumed rapid caldera uplift caused by trapdoor faulting within Kita-Ioto Caldera (Figure 3).
- Distinct fault rupture locations: The tsunami waveforms from the 2017 and 2019 earthquakes displayed significant differences. These discrepancies are well explained by differences in the rupture locations within the caldera fault system; the 2017 earthquake was associated with the western fault (Figure 3a), while the 2019 earthquake corresponded to the northern fault (Figure 3b).

Significance of this study
- Detection of micro-tsunamis and application to submarine volcano studies:Applying the waveform stacking method, we successfully detected high-quality tsunami waveforms with 1–2 mm amplitudes in a tsunami observation network located approximately 1,000 km from the earthquake source. This advancement enables detailed investigations into the volcanic activity of the Kita-Ioto Caldera.
- Evidence for repeated trapdoor faulting: Building on our previous study (Sandanbata & Saito, 2024a) that analyzed the 2008 and 2015 earthquakes, our analysis of the micro-tsunami records from the 2017 and 2019 earthquakes provides further robust evidence that trapdoor faulting recurs every 2–5 years in Kita-Ioto Caldera, causing recurrent large-scale caldera uplifts.
- Segmentation of intra-caldera faults:Our findings that the 2017 and 2019 trapdoor faulting events occurred on different fault segments imply that multiple fault segments rupture alternately in Kita-Ioto Caldera. This fault segmentation may cause the high-frequency recurrence of trapdoor faulting at intervals of only a few years (Figure 4).

Future Perspectives
Future studies should focus on the detailed mechanisms of these repeated faulting events and their relationship with magma accumulation processes beneath the caldera. Such efforts will potentially improve our ability to predict the long-term volcanic activity of Kita-Ioto Caldera and contribute to assessing future risks of submarine eruptions.
References:
Sandanbata, O., & Saito, T. (2024a). Quantifying Magma Overpressure Beneath a Submarine Caldera: A Mechanical Modeling Approach to Tsunamigenic Trapdoor Faulting Near Kita-Ioto Island, Japan. Journal of Geophysical Research, [Solid Earth], 129(1), e2023JB027917. https://doi.org/10.1029/2023JB027917
Sandanbata, O., Watada, S., Satake, K., Kanamori, H., Rivera, L., & Zhan, Z. (2022). Sub‐decadal volcanic tsunamis due to submarine trapdoor faulting at Sumisu caldera in the Izu–Bonin arc. Journal of Geophysical Research, [Solid Earth], 127(9), e2022JB024213. https://doi.org/10.1029/2022jb024213