S51D-2729
A subsurface structure change associated with the eruptive activity at Sakurajima Volcano, Japan, inferred from an accurately controlled source

Friday, 18 December 2015
Poster Hall (Moscone South)
Yuta Maeda1, Koshun Yamaoka1, Hiroki Miyamachi2, Toshiki Watanabe3, Takahiro Kunitomo1, Ryoya Ikuta4, Hiroshi Yakiwara5 and Masato Iguchi6, (1)Nagoya University, Nagoya, Japan, (2)Kagoshima University, Kagoshima, Japan, (3)Earthquake Research Institute, University of Tokyo, Tokyo, Japan, (4)Shizuoka University, Shizuoka, Japan, (5)Nansei-toko Observatory, Kagoshima, Japan, (6)Sakurajima Volcanic Observatory, Kagoshima, Japan
Abstract:
Temporal variations of Green functions associated with the eruptive activity at Sakurajima Volcano, Japan, were estimated using an accurately controlled routinely operated signal system (ACROSS). We used the seismic records at stations HAR and KAB, which are 615 m and 1200 m apart from the ACROSS source, respectively, that recorded a modulating sinusoidal signal in a 12.505 ± 2.5 Hz band from the ACROSS. We deconvolved individual 400 s records at these stations by a known source time function and stacked the results based on the time relative to individual eruptions and eruption intervals; the quantities obtained by this procedure are Green functions corresponding to various stages of the eruptive activity. At station HAR, we identified an energy density decrease in a 2–4 s window of the Green functions before and after individual eruptions and for time periods of short eruption intervals. This time window (2–4 s) is after the main phase and before an exponentially decaying coda, and dominated by a later phase of a flat amplitude. At station KAB, similar energy density decrease was observed in a 2–6 s window. The Green functions estimated by our procedure may be affected by volcanic earthquakes. We evaluated this effect using seismic records in a time span when the ACROSS was temporarily not operating. This evaluation showed that an apparent Green function change caused by the volcanic earthquakes would be spread out over the entire 50 s window of the Green functions. Therefore the Green function change localized in the later phase portion (2–4 s at HAR and 2–6 s at KAB) is difficult to explain by contamination from volcanic earthquakes and tremors. The change could be more reasonably attributed to a subsurface structure change caused by the volcanic activity.