Magnetotelluric study on a vapor-dominated geothermal reservoir in the Matsukawa area, Japan

•Three-dimensional magnetotelluric inversion was conducted in the Matsukawa geothermal area, northern Japan.•Estimated resistivity values agreed with the resistivity logging data.•Low permeability shielding, which has been considered important for the formation of vapor-dominated reservoirs, was captured as a high resistivity structure.•Using the logging data and the resistivity values obtained from magnetotelluric inversion, temperature estimation was performed by the neural kriging method. Vapor-dominated geothermal areas produce only vapor from geothermal reservoirs and have been reported worldwide (e.g., Larderello, Italy; The Geysers, United States; and Kamojang, Indonesia) and studied from several perspectives. White et al. (1971) proposed a typical vapor-dominated reservoir model, which has been applied to numerous vapor-dominated fields. The Matsukawa area, located in Iwate prefecture, is one of the few vapor-dominated geothermal areas in Japan that contains an operational power plant; therefore, it has been the site of many geological and drilling studies. For example, Ozeki et al. (2001) proposed a geological conceptual model of this area. In this study, we investigated the characteristic resistivity structure in a vapor-dominated geothermal reservoir in the Matsukawa geothermal area. We performed a three-dimensional (3D) inversion with ModEM (Egbert and Kelbert, 2012) for magnetotelluric data obtained from 68 stations. We determined an optimized solution according to the results with the smallest root mean square value by considering various calculation conditions, including the smoothing parameter and Lagrange multiplier. The 3D inversion results were used to confirm clay alteration zones, faulting, basement rocks, and low-permeability lateral shielding, which characterizes the vapor-dominated geothermal system in this area. The 3D resistivity structure is considered valid because it is consistent with logging data. In addition, we estimated the 3D temperature structure using the neural kriging method (e.g., Ishitsuka et al., 2018) based on the obtained resistivity structure, temperature logging data, and electrical logging data.