5:15 PM - 6:30 PM
[SVC28-P17] Two experiments using a small-aperture infrasonic array at Kirishima
Keywords:Infrasound, array, volcano
Infrasound, which is the air vibration with sub-audible frequency (< 20 Hz), is generated by various geophysical phenomena and is used for monitoring them. A small-aperture (10-30 m) infrasonic array with high-performance microphones is shown to have a practical resolution of the direction of arrival of an incident signal [Johnson & Palma, 2015; Yamakawa et al., 2018]. Also, a new infrasonic MEMS (Micro Electro Mechanical Systems) microphone is being developed, which is inexpensive and has a reasonably good performance [Wada and Takahashi, Japanese Journal of Applied Physics, 2020]. In this study, we conducted two experiments using a small aperture array at Kirishima volcano in 2020.
The first experiment examined the performance of the MEMS microphones [Wada and Takahashi, 2020]. It was conducted for 19 hours from 5 p.m. on 19 October 2020 at Kirishima Ioyama, where several fumaroles were active. The MEMS microphones were compared with the latest model of infrasound microphones (Type 7744N, ACO CO., LSD., hereafter referred to as ACO microphone) that had been used in several infrasonic studies. We installed two identical three-element infrasonic arrays with an aperture of 15 meters (Figure 1). The one consisted of the MEMS microphones and the other of the ACO microphones. At each array element, a MEMS microphone and an ACO microphone were put in the same box. There were two active fumaroles around the array. One was Iwo-yama West Craters, 150 m northeast of the array, and the other was Iwo-yama South Crater, 270 m northwest of the array [Tajima et al., 2020; Muramatsu et al., 2021]. Here we refer to the fumarole of Iwo-yama West Craters as NW fumarole and the fumarole of Iwo-yama South Crater as NE fumarole. We also placed a microphone (Type SI104, Hakusan Co., hereafter referred to as SI microphone) near the NW fumarole.
The recorded waveforms and the array processing results were excellently matched between the MEMS microphone array and the ACO microphone array, confirming that the MEMS microphones are practically useful.
The array processing showed that the infrasound from the NE fumarole was dominant at the arrays. We found that the NW fumarole also emanated infrasound by the combined analysis of the array data with the data from the SI microphone located close to the NW fumarole. The NW signals were weak compared to the NE signals: the cross-correlation coefficient between the array microphones and the SI microphone was 0.13 for the NE and 0.01 for the NW (Figure 2). With the monopole source assumption and the ratio of the correlation coefficient, the NW-to-NE amplitude ratio was approximately 0.06-0.1.
The second experiment was a mobile array survey for infrasound noise. It was conducted around Makizono-cho, Kirishima city, from 3 p.m. to 5 p.m. on 20 October 2020. While traveling by car, we stopped at an observation point, deployed the small-aperture (15 m) array for several minutes, retrieved them, and then moved to the next point. We used the ACO microphones that were used in the first experiment. We obtained the infrasonic array data at four locations in 2 hours.
The results indicated that the dominant infrasound noise was from the fumarolic area (Iodani Fumarolic Area Park) at the P1, P2, and P4 (Figure 3). At the P3, the estimated source direction was deviated. This deviation should have been brought by the topographical barrier of a hill between P3 and the fumarolic area. Fumarolic noise is persistent and can disturb infrasonic monitoring of volcanoes. The array survey for background noise will be useful for designing infrasound observations for volcanoes.
Two experiments were successful. In the first experiment, new MEMS microphones was proven as practical. This new and inexpensive microphone will benefit the infrasonic studies. The first experiment also succeeded in detecting the existence of the infrasound signals that were masked by another signals. In the second experiment, we proved the movability of a small infrasonic array, and succeeded to identify the infrasonic noise source in a short time. These two experiments demonstrated the wide usage of a small infrasonic array.
The first experiment examined the performance of the MEMS microphones [Wada and Takahashi, 2020]. It was conducted for 19 hours from 5 p.m. on 19 October 2020 at Kirishima Ioyama, where several fumaroles were active. The MEMS microphones were compared with the latest model of infrasound microphones (Type 7744N, ACO CO., LSD., hereafter referred to as ACO microphone) that had been used in several infrasonic studies. We installed two identical three-element infrasonic arrays with an aperture of 15 meters (Figure 1). The one consisted of the MEMS microphones and the other of the ACO microphones. At each array element, a MEMS microphone and an ACO microphone were put in the same box. There were two active fumaroles around the array. One was Iwo-yama West Craters, 150 m northeast of the array, and the other was Iwo-yama South Crater, 270 m northwest of the array [Tajima et al., 2020; Muramatsu et al., 2021]. Here we refer to the fumarole of Iwo-yama West Craters as NW fumarole and the fumarole of Iwo-yama South Crater as NE fumarole. We also placed a microphone (Type SI104, Hakusan Co., hereafter referred to as SI microphone) near the NW fumarole.
The recorded waveforms and the array processing results were excellently matched between the MEMS microphone array and the ACO microphone array, confirming that the MEMS microphones are practically useful.
The array processing showed that the infrasound from the NE fumarole was dominant at the arrays. We found that the NW fumarole also emanated infrasound by the combined analysis of the array data with the data from the SI microphone located close to the NW fumarole. The NW signals were weak compared to the NE signals: the cross-correlation coefficient between the array microphones and the SI microphone was 0.13 for the NE and 0.01 for the NW (Figure 2). With the monopole source assumption and the ratio of the correlation coefficient, the NW-to-NE amplitude ratio was approximately 0.06-0.1.
The second experiment was a mobile array survey for infrasound noise. It was conducted around Makizono-cho, Kirishima city, from 3 p.m. to 5 p.m. on 20 October 2020. While traveling by car, we stopped at an observation point, deployed the small-aperture (15 m) array for several minutes, retrieved them, and then moved to the next point. We used the ACO microphones that were used in the first experiment. We obtained the infrasonic array data at four locations in 2 hours.
The results indicated that the dominant infrasound noise was from the fumarolic area (Iodani Fumarolic Area Park) at the P1, P2, and P4 (Figure 3). At the P3, the estimated source direction was deviated. This deviation should have been brought by the topographical barrier of a hill between P3 and the fumarolic area. Fumarolic noise is persistent and can disturb infrasonic monitoring of volcanoes. The array survey for background noise will be useful for designing infrasound observations for volcanoes.
Two experiments were successful. In the first experiment, new MEMS microphones was proven as practical. This new and inexpensive microphone will benefit the infrasonic studies. The first experiment also succeeded in detecting the existence of the infrasound signals that were masked by another signals. In the second experiment, we proved the movability of a small infrasonic array, and succeeded to identify the infrasonic noise source in a short time. These two experiments demonstrated the wide usage of a small infrasonic array.