9:45 AM - 10:00 AM
[SEM14-04] Toward joint inversion of AMT and ACTIVE data sets for better understanding of the hydrothermal system of Aso volcano
Keywords:inversion, resistivity, AMT, volcano, Aso, CSEM
The resistivity structure beneath a volcano is an important proxy for understanding hydrothermal and magmatic systems. The Audio-frequency Magnetotelluric (AMT) survey, which utilizes magnetic variations due to natural ionospheric disturbances and lightning, is a popular method used to investigate shallow resistivity structures. On the other hand, controlled-source electromagnetic (CSEM) methods are also important mainly for the monitoring purpose, because of their stable signal-to-noise ratio. A CSEM volcano monitoring method called ACTIVE (Array of Controlled Transient-electromagnetics for Imaging Volcano Edifice; Utada et al. 2007) can reveal the temporal change in resistivity structures. Minami et al. (2018) applied the three-dimensional (3-D) finite element inversion to the ACTIVE data sets and inferred the temporal variation in the 3-D resistivity structure associated with the magmatic eruption of Aso volcano, Japan, starting in November 2014.
However, to date, there are no effective inversion tools to infer a resistivity structure using AMT and ACTIVE data sets simultaneously. To fill this gap, we are developing an inversion code that can jointly invert AMT (e.g. Kanda et al. 2008) and ACTIVE data sets (e.g. Minami et al. 2018). Since ACTIVE consists of an array of induction-coil receivers that measure the vertical component of the magnetic variation and grounded electric dipole sources, ACTIVE data sets are theoretically free from the effect of galvanic distortion. This is also an advantage of using both AMT and ACTIVE data sets in the same inversion. Furthermore, the AMT-ACTIVE joint inversion can use a common coefficient matrix in solving each linear problem when the Dirichlet boundary conditions are adopted in both ACTIVE and MT problems. This means that the ACTIVE inversion code with a direct solver and the Dirichlet boundary condition developed by Minami et al. (2018) is easily extended to the joint inversion with AMT data sets without modification of the coefficient matrix. In the presentation, we report the algorithm of our joint inversion code and show preliminary results of joint inversions of AMT and ACTIVE data sets obtained at Aso volcano.
However, to date, there are no effective inversion tools to infer a resistivity structure using AMT and ACTIVE data sets simultaneously. To fill this gap, we are developing an inversion code that can jointly invert AMT (e.g. Kanda et al. 2008) and ACTIVE data sets (e.g. Minami et al. 2018). Since ACTIVE consists of an array of induction-coil receivers that measure the vertical component of the magnetic variation and grounded electric dipole sources, ACTIVE data sets are theoretically free from the effect of galvanic distortion. This is also an advantage of using both AMT and ACTIVE data sets in the same inversion. Furthermore, the AMT-ACTIVE joint inversion can use a common coefficient matrix in solving each linear problem when the Dirichlet boundary conditions are adopted in both ACTIVE and MT problems. This means that the ACTIVE inversion code with a direct solver and the Dirichlet boundary condition developed by Minami et al. (2018) is easily extended to the joint inversion with AMT data sets without modification of the coefficient matrix. In the presentation, we report the algorithm of our joint inversion code and show preliminary results of joint inversions of AMT and ACTIVE data sets obtained at Aso volcano.