3:45 PM - 4:00 PM
[STT37-02] Three-dimensional imaging of remagnetization beneath Tarumae Volcano by repetitive aeromagnetic survey using an unmanned helicopter
Widespread use of UAVs has led to an increased number of aeromagnetic surveying by UAVs in volcanic regions (e.g., Kaneko et al., 2011; Koyama et al., 2013; Shibuya et al., 2021; Tada et al., 2021). The authors’ research group has been conducting volcano surveys using uncrewed helicopter system in cooperation with the Hokkaido Regional Development Bureau. The primary purpose of the project is visual inspection, various measurements, and material sampling from outside the danger zone when the volcanic activity is elevated. In this context, the authors have conducted aeromagnetic surveys in 2011, 2012, 2013, 2020, and 2021 at Mt. Tarumae intending to detect temporalchanges associated with volcanic activity.
Shibuya et al. (2021) analyzed the survey data of 2013 and 2020 and succeeded in detecting a remarkable temporal change for the first time in this project. They revealed that the magnetic moment increased right beneath the summit lava dome. Meanwhile, the activity of the volcano generally stayed calm and did not change significantly during this period, although some background fumarolic and seismic activities accompanied it. This study re-analyzed the same temporal magnetic change using the different methods described below. In order to understand the cause of remagnetization, we visualized the source distribution and compared it with the hypocenters of micro-earthquakes and the resistivity structure.
We performed two types of inversion: a model assuming a single magnetic dipole source (the dipole model) and a model with a set of uniformly magnetized cuboids and obtaining the 3D distribution of the magnetization change (the block model). For the dipole model, we used an MCMC inversion to evaluate the uncertainty of the solution. On the other hand, we used the L1-L2 norm 3D inversion developed by Utsugi (2019) for the block model, which allows models with sharp boundaries and is effective for a potentially heterogeneous target such as volcanoes.
We obtained the optimal source right beneath the lava dome in the dipole model. The posterior probability distributions of the model parameters showed that the variance in the vertical source location was about twice as large as in the horizontal ones. We also recognized a significant trade-off between source intensity and depth. Nonetheless, the 95% confidence range for depth was less than 200 m. On the other hand, the block model produced a columnar zone of remagnetization extending upward from the bottom side of the optimal dipole source to the ground surface, slightly bending toward Crater-A (Figs. 1 and 2). The total remagnetized moment well matched that of the optimal single dipole, suggesting that the two inversion results were consistent.
The magnetized zone in the block model overlapped with the micro-earthquakes beneath the lava dome, which was relocated by Tsuchiya (2010) using the Double Difference method. The source region was roofed by the uppermostpart of the low resistivity zone, which Yamaya et al. (2009) interpreted as an impermeable clay layer (Fig. 2). In addition, the Japan Meteorological Agency reported that the fumarolic temperature of Crater-A gradually decreased from ca. 600 to 500℃ during the period. Based on these facts, it is likely that the remagnetized zone corresponds to a conduit through which hot gases ascend. Namely, we consider that the depleted supply of the gases cooled the conduit and made it remagnetized along the present geomagnetic field. Therefore, thermal deformation and associated cracking in and around the conduit may have caused the micro-seismicity.
Acknowledgments: We obtained the aeromagnetic data in this study with the cooperation of the Hokkaido Development Bureau and the Earthquake Research Institute, the University of Tokyo.
Fig.1 (left) The remagnetized region obtained from the block model of this study.
Fig.2 (right) Comparison with other sources in the EW cross-section.
Shibuya et al. (2021) analyzed the survey data of 2013 and 2020 and succeeded in detecting a remarkable temporal change for the first time in this project. They revealed that the magnetic moment increased right beneath the summit lava dome. Meanwhile, the activity of the volcano generally stayed calm and did not change significantly during this period, although some background fumarolic and seismic activities accompanied it. This study re-analyzed the same temporal magnetic change using the different methods described below. In order to understand the cause of remagnetization, we visualized the source distribution and compared it with the hypocenters of micro-earthquakes and the resistivity structure.
We performed two types of inversion: a model assuming a single magnetic dipole source (the dipole model) and a model with a set of uniformly magnetized cuboids and obtaining the 3D distribution of the magnetization change (the block model). For the dipole model, we used an MCMC inversion to evaluate the uncertainty of the solution. On the other hand, we used the L1-L2 norm 3D inversion developed by Utsugi (2019) for the block model, which allows models with sharp boundaries and is effective for a potentially heterogeneous target such as volcanoes.
We obtained the optimal source right beneath the lava dome in the dipole model. The posterior probability distributions of the model parameters showed that the variance in the vertical source location was about twice as large as in the horizontal ones. We also recognized a significant trade-off between source intensity and depth. Nonetheless, the 95% confidence range for depth was less than 200 m. On the other hand, the block model produced a columnar zone of remagnetization extending upward from the bottom side of the optimal dipole source to the ground surface, slightly bending toward Crater-A (Figs. 1 and 2). The total remagnetized moment well matched that of the optimal single dipole, suggesting that the two inversion results were consistent.
The magnetized zone in the block model overlapped with the micro-earthquakes beneath the lava dome, which was relocated by Tsuchiya (2010) using the Double Difference method. The source region was roofed by the uppermostpart of the low resistivity zone, which Yamaya et al. (2009) interpreted as an impermeable clay layer (Fig. 2). In addition, the Japan Meteorological Agency reported that the fumarolic temperature of Crater-A gradually decreased from ca. 600 to 500℃ during the period. Based on these facts, it is likely that the remagnetized zone corresponds to a conduit through which hot gases ascend. Namely, we consider that the depleted supply of the gases cooled the conduit and made it remagnetized along the present geomagnetic field. Therefore, thermal deformation and associated cracking in and around the conduit may have caused the micro-seismicity.
Acknowledgments: We obtained the aeromagnetic data in this study with the cooperation of the Hokkaido Development Bureau and the Earthquake Research Institute, the University of Tokyo.
Fig.1 (left) The remagnetized region obtained from the block model of this study.
Fig.2 (right) Comparison with other sources in the EW cross-section.