- 1Tohoku University, Sendai, Japan
- 2Nagoya University, Nagoya, Japan
- 3Kyoto University, Uji, Japan
- 4Kyushu University, Fukuoka, Japan
Seismic shaking introduces dynamic strain/stress changes to volcanic edifices and could activate volcanic systems. Continuous monitoring of crustal seismic velocity using seismic interferometry is a powerful approach to detect stress changes and dynamic magmatic processes that are otherwise difficult to observe. Applying this technique to DAS data with thousands of channels enables unprecedented spatial resolution for subsurface structure monitoring.
We analyzed continuous DAS data between September 2024 and October 2025 at Sakurajima volcano, Japan, using the optical fiber cable installed along a loop road around the volcano. Seismic ambient noise cross-correlation functions (CCFs) were computed for approximately 1.22 million channel pairs. The AOBA-S high-performance computing system at Tohoku University Cyberscience Center allowed us to drastically reduce the estimated computation time from 46 years to 26 days. Relative velocity changes were measured by applying the doublet method to 20-day stacked CCFs in the 0.25–0.5 Hz band.
Significant coseismic velocity decreases and velocity recovery were detected. After the Mw 6.8 Hyuganada earthquake in January 2025, we observed a velocity decrease of 0.09%. Tomographic analysis revealed a pronounced velocity decrease (0.12%) around the crater. Smaller changes (0.01%) occurred during the Mw 6.0 Osumi earthquake in April 2025, and recovery after the Mw 7.0 Hyuganada earthquake in August 2024 was also identified. Velocity recovery followed a logarithmic trend (dv/v = m log₁₀t + A), with the recovery rate m decreasing from 0.173 after the Mw 7.0 event to 0.055 and 0.030 after the subsequent Mw 6.8 and Mw 6.0 events, respectively. The logarithmic recovery is consistent with relaxation of cracks with different aspect ratios, and the decreasing m values suggest progressive exhaustion of easily healed cracks through successive events [Illien et al., 2025]. Moreover, paths crossing the crater area showed faster recovery than peripheral paths, possibly reflecting elevated temperatures and/or the presence of fluids near the crater [Brantut, 2015; Snieder et al., 2017].
These results demonstrate that DAS-based seismic interferometry can resolve spatiotemporal velocity changes with exceptional detail, offering new insights into the response of volcanic systems to seismic shaking.
Acknowledgments: We used fiber optic cables of the Ministry of Land, Infrastructure, Transport and Tourism. We would like to thank the Osumi River National Highway Office for helping us with the DAS observation. We also thank the Tohoku University Cyberscience Center for providing access to the AOBA-S high-performance computing system for CCF calculations.
How to cite: Hirose, T., Nishimura, T., Nakahara, H., Shimomura, Y., Takizawa, H., Taguchi, K., Nakamichi, H., Emoto, K., Yonemori, K., and Abdul Rahman, S. I.: Coseismic velocity decreases and logarithmic recovery at Sakurajima volcano imaged by DAS-based seismic interferometry, Galileo conference: Fibre Optic Sensing in Geosciences, Aussois, France, 31 Aug–4 Sep 2026, GC14-FibreOptic-5, https://doi.org/10.5194/egusphere-gc14-fibreoptic-5, 2026.