The magnetotelluric (MT) method is a well-established tool in geothermal exploration. Case studies from all over the world and from different geothermal settings have proven its effectiveness, when it comes to subsurface reservoir characterization and the successful siting of geothermal wells. A reason for MT being a popular tool in geothermal exploration is that the bulk electrical conductivity of the subsurface, as recovered by MT, can be used as a proxy for key geothermal parameters. For example, fluid saturation and connectedness, hydrothermal alteration and active magmatic heat sources all significantly influence electrical conductivity and appear as electrically conducting zones in the subsurface. In the field, MT surveying benefits from little manpower requirements, low environmental impact and from the fact that natural electromagnetic source signals are permanently present everywhere on the globe. Whilst MT is successfully used in sparsely populated regions, challenges arise when it comes to MT exploration in populated areas. Here, data acquisition is prone to noise issues that arise from local infrastructure that make MT often too cumbersome for commercial applications. The interpretation of MT data is improving continuously with the development of powerful numerical modelling tools. 3-D subsurface models with flexible meshes that adapt to topography and varying data resolution allow one to characterize geothermal systems from their surface manifestations down to their deeper roots in the lower crust.

We present case studies from three high-temperature geothermal systems in the East African Rift, where MT is successfully used to image magmatic reservoirs that drive convection of hydrothermal fluids. As demonstrated, recovered MT models of these systems play a key role when it comes to the successful siting of geothermal wells, which are commonly drilled into permeable up-flow zones above shallow magmatic reservoirs. In another case study from an intermediate-temperature geothermal system in the Mongolian Hangai, we demonstrate how large datasets with more than 300 MT stations can be acquired by small academic teams. However, the interpretation of MT subsurface models from such intermediate-temperature systems in non-volcanic terrains is less straight forward and requires more a priori knowledge and interdisciplinary strategies as compared to MT studies of high-temperature volcanic geothermal systems.