Magnetotelluric (MT) sounding is a popular choice of a geophysical method for geothermal exploration (e.g. Newman et al., 2008). Geoelectrical model of the subsurface can be recovered by 3D inversion of MT data. Distribution of electrical conductivity obtained from MT inversion can help identify altered areas saturated with hydrothermal fluids, indicate possible fluid paths, as well as locate deep reservoir itself.
In our paper we derive an efficient method for inversion of MT data using Gauss-Newton method in data space. Gauss-Newton method is one of the most popular approaches to solving inverse problems in geophysics. The major obstacle to applying the Gauss-Newton method in model space is the need to invert a large square matrix with both dimensions equal to the number of model parameters. In modern applications, the number of model parameters can be on the order of millions. Storage and operations with a matrix of this size can be unmanageable on a typical workstation, or even on a moderately sized cluster. Several researchers (e.g. Kordy et al., 2016) described and applied Gauss-Newton inversion in the data space. The size of the corresponding square Hessian matrix in the data space is reduced to both dimensions equal to the number of data points. The data space RGN formulation is equivalent to the model space formulation.
An example of inversion of MT data collected in the area with known geothermal resources is presented. We applied our method to the data from Sevier Thermal Belt located in southwestern Utah. Resulting conductivity model indicated possible location of the magma reservoir, possible paths of geothermal fluids, and the location of saturated altered rocks in the subsurface.
Kordy, M., P.Wannamaker, V. Maris, E. Cherkaev, and G. Hill, 2016, 3-dimensional magnetotelluric inversion including topography using deformed hexahedral edge finite elements and direct solvers parallelized on symmetric multiprocessor computers – Part II: direct data-space inverse solution, Geophysical Journal International, 204, 94-110.
AAPG Datapages/Search and Discovery Article #90323 ©2018 AAPG Annual Convention and Exhibition, Salt Lake City, Utah, May 20-23, 2018