5:15 PM - 7:15 PM
[SEM15-P05] Estimation of underground resistivity structure of Kozushima using tsunami-induced electric fields during the 2011 Tohoku tsunam
Keywords:Electric field, tsunami, resistivity structure, Kozushima, the 2011 Tohoku earthquake
When highly conductive seawater moves through the geomagnetic field due to a tsunami, tsunami-generated electromagnetic (TGEM) field variations occur (e.g., Tyler, 2005, GRL). The TGEM variations are observable both on land and on the seafloor, as recorded during the 2011 Tohoku tsunami (e.g., Minami et al., 2017, JGR). On Kozushima, one of the Izu Islands, tsunami-generated electric (TGE) variations were observed almost simultaneously with the peak tide gauge measurement during the 2011 Tohoku tsunami (Nakatani, 2015, Graduation thesis in Tokyo Gakugei University). While TGE variations observed on the seafloor are minimally affected by the subseafloor resistivity (Shimizu and Utada, 2015, GJI), observed on islands are strongly influenced by the underground resistivity (Minami, 2024, PTA; Shibahara, 2022, Graduation thesis in Kobe University). It has been reported that variations observed on islands are larger when the underground resistivity is higher. In this study, we attempt to estimate the underground resistivity structure of Kozushima using the TGE variation data during the 2011 Tohoku tsunami.
On Kozushima, electric potential observations utilizing two long dipoles were conducted by Dr. Yoshiaki Orihara and Mr. Yoichi Noda as part of the “Earthquake International Frontier Research” initiative of the Science and Technology Agency. One dipole oriented northeast-southwest (2.137km), and the other oriented southeast-northwest (2.382km). We converted the potential difference data into electric field data for the north-south and east-west components. The tidal records of Kozushima were obtained from tide gauges operated by the Japan Coast Guard. We simulated TGE variations with different underground resistivity structures. In the electric field simulation on Kozushima, we used a tsunami velocity field generated by the tsunami simulation code COMCOT (Wang and Liu, 2006, JHR), which was then input into the time-domain finite-element TGEM simulation code TMTGEM (Minami et al., 2017, JGR). The tsunami source was based on fault model proposed by Satake et al., (2013, BSSA). For the underground resistivity structure of Kozushima, we considered homogeneous models with resistivities of 1000, 100, 10 Ωm, as well as models with low-resistivity regions located along the eastern or western coastal areas of Kozushima, or surrounding the entire coastline. Furthermore, to accurately compare with the observed data, pseudo-observation points were placed along the dipoles of potential difference observation. The TGEM simulations at these pseudo-observation points were integrated along each dipole direction to compute potential difference data.
In the TGE variations with homogeneous resistivity models of Kozushima, the amplitude of the variations increased as the resistivity value increased. Comparing the polarities of TGE fields, the observations showed dominant east-west variations, whereas the homogeneous model exhibited dominant north-south variations, failing to reproduce the observed data. Assuming the low-resistivity regions along the eastern coast or surrounding the entire coastline produced variations in the southeast direction, closely matching the observations in both direction and amplitude. In contrast, when low-resistivity regions were assumed along the western coast, the results were similar to the homogeneous models, with dominant variations in the north-south direction.
These results indicate the presence of low-resistivity regions along the eastern coast of Kozushima. Since the low-resistivity region along the western coast produced results similar to those of the homogeneous model—and given that the tsunami approached Kozushima from the east—the resistivity structure on the western side likely has a small influence, making it difficult to estimate. In this study, we plan to further refine our resistivity structure estimates to better matche the observations and further investigate the sensitivity of TGE fields to the resistivity structure on Kozushima in more detail
On Kozushima, electric potential observations utilizing two long dipoles were conducted by Dr. Yoshiaki Orihara and Mr. Yoichi Noda as part of the “Earthquake International Frontier Research” initiative of the Science and Technology Agency. One dipole oriented northeast-southwest (2.137km), and the other oriented southeast-northwest (2.382km). We converted the potential difference data into electric field data for the north-south and east-west components. The tidal records of Kozushima were obtained from tide gauges operated by the Japan Coast Guard. We simulated TGE variations with different underground resistivity structures. In the electric field simulation on Kozushima, we used a tsunami velocity field generated by the tsunami simulation code COMCOT (Wang and Liu, 2006, JHR), which was then input into the time-domain finite-element TGEM simulation code TMTGEM (Minami et al., 2017, JGR). The tsunami source was based on fault model proposed by Satake et al., (2013, BSSA). For the underground resistivity structure of Kozushima, we considered homogeneous models with resistivities of 1000, 100, 10 Ωm, as well as models with low-resistivity regions located along the eastern or western coastal areas of Kozushima, or surrounding the entire coastline. Furthermore, to accurately compare with the observed data, pseudo-observation points were placed along the dipoles of potential difference observation. The TGEM simulations at these pseudo-observation points were integrated along each dipole direction to compute potential difference data.
In the TGE variations with homogeneous resistivity models of Kozushima, the amplitude of the variations increased as the resistivity value increased. Comparing the polarities of TGE fields, the observations showed dominant east-west variations, whereas the homogeneous model exhibited dominant north-south variations, failing to reproduce the observed data. Assuming the low-resistivity regions along the eastern coast or surrounding the entire coastline produced variations in the southeast direction, closely matching the observations in both direction and amplitude. In contrast, when low-resistivity regions were assumed along the western coast, the results were similar to the homogeneous models, with dominant variations in the north-south direction.
These results indicate the presence of low-resistivity regions along the eastern coast of Kozushima. Since the low-resistivity region along the western coast produced results similar to those of the homogeneous model—and given that the tsunami approached Kozushima from the east—the resistivity structure on the western side likely has a small influence, making it difficult to estimate. In this study, we plan to further refine our resistivity structure estimates to better matche the observations and further investigate the sensitivity of TGE fields to the resistivity structure on Kozushima in more detail