5:15 PM - 7:15 PM
[SVC32-P12] Temporal variation of volcanic Long-Period events during the 2018 unrest at Kusatsu-shirane
Keywords:Volcano seismology, Volcanic Long-Period event, Volcano monitoring
Volcanic Long-Period events (LP events) have repeatedly occurred at Kurastu-shirane. LP events are characterized by frequencies (f) and a quality factor (Q) of damped oscillations, which are thought to be generated in a fluid-filled resonator. At Kurastu-shirane, based on estimates of crack geometries and fluid properties, LP events are considered as resonance oscillations triggered by volumetric changes associated with condensation of water vapor in a crack (e.g. Taguchi et al., JGR, 2018; Nakano et al., 2025). Nakano et al. (2025) showed that estimates of water vapor volume in a crack are correlated with the inverse of the lowest spectral peak frequency (f1). These results suggest that f1 is a useful parameter for monitoring the hydrothermal activity at Kusatsu-shirane. However, previous studies focused only on LP events during 1989-1993. Thus, it is unclear whether the above correlation holds during different periods. At Kusatu-shirane, many LP events have also been observed since 2018. For these recent observations, a comparison of seismic data with the tiltmeter data can be performed. Investigating these recent events may provide new insights into LP source processes. Therefore, we estimated source locations, f1, and Q, and investigated their temporal changes for LP events during 2018.
In this study, we focused on LP events occurring between February and July in 2018. We used vertical seismic waveform data recorded at 9 seismic stations operated by Institute of Science Tokyo and Japan Meteorological Agency at Kusatu-shirane. We estimated source locations of the LP events based on picked first-arrival times. For each LP event waveform at each station, we calculated an amplitude spectrum and picked f1. Next, we applied a band-pass filter (f1 ± 0.5 Hz) to the seismic waveform and obtained the maximum amplitude (A0). The Q value was estimated by fitting the damped oscillation part of the filtered waveform to the theoretical equation by Maeda et al. (JGR, 2013). We calculated f1, A0, and Q for each event as averages across all stations.
Our estimated sources were located around the crater, although they were scattered. Their distributions were consistent with the cluster of high-frequency seismic events (HF) that occurred beneath the Yugama crater. Some LP events were located near the sub-horizontal cracks estimated as the LP sources by previous studies (e.g. Nakano & Kumagai, GRL, 2005), but many events also occurred at deeper positions. Further investigations are needed to assess the accuracy of the source distribution. We found f1 fluctuates between 1.6 and 3.4Hz except for 3 events. However, f1 showed an upward trend of f1 when considering entire analysis period. Q and A0 were scattered. Notably, the f1 time series between May and July showed systematic U- or V-shaped patterns, featuring decreasing phases lasting roughly two to three days and increasing phases of approximately two weeks. As f1 continuously decreased, intervals of LP event shortened, while the daily number of HF events slightly increased. LP events during 1989 and 1993 also showed systematic temporal changes in f1 (Nakano et al., 2025), resembling the patterns we found. This suggests that common triggering processes exist for LP events in both 2018 and 1989-1993. In future studies, we need to investigate the relationships among LP events, high-frequency seismic events and fumarole activity to further understand the shallow thermal system at Kuratsu-shirane.
In this study, we focused on LP events occurring between February and July in 2018. We used vertical seismic waveform data recorded at 9 seismic stations operated by Institute of Science Tokyo and Japan Meteorological Agency at Kusatu-shirane. We estimated source locations of the LP events based on picked first-arrival times. For each LP event waveform at each station, we calculated an amplitude spectrum and picked f1. Next, we applied a band-pass filter (f1 ± 0.5 Hz) to the seismic waveform and obtained the maximum amplitude (A0). The Q value was estimated by fitting the damped oscillation part of the filtered waveform to the theoretical equation by Maeda et al. (JGR, 2013). We calculated f1, A0, and Q for each event as averages across all stations.
Our estimated sources were located around the crater, although they were scattered. Their distributions were consistent with the cluster of high-frequency seismic events (HF) that occurred beneath the Yugama crater. Some LP events were located near the sub-horizontal cracks estimated as the LP sources by previous studies (e.g. Nakano & Kumagai, GRL, 2005), but many events also occurred at deeper positions. Further investigations are needed to assess the accuracy of the source distribution. We found f1 fluctuates between 1.6 and 3.4Hz except for 3 events. However, f1 showed an upward trend of f1 when considering entire analysis period. Q and A0 were scattered. Notably, the f1 time series between May and July showed systematic U- or V-shaped patterns, featuring decreasing phases lasting roughly two to three days and increasing phases of approximately two weeks. As f1 continuously decreased, intervals of LP event shortened, while the daily number of HF events slightly increased. LP events during 1989 and 1993 also showed systematic temporal changes in f1 (Nakano et al., 2025), resembling the patterns we found. This suggests that common triggering processes exist for LP events in both 2018 and 1989-1993. In future studies, we need to investigate the relationships among LP events, high-frequency seismic events and fumarole activity to further understand the shallow thermal system at Kuratsu-shirane.