Development of a Novel Parameterisation for Three-Dimensional Magnetotelluric Inversion of Continental Lithosphere

This study develops an original parameterisation framework for three-dimensional magnetotelluric inversion that improves model resolution and geologically meaningful parameter recovery for continental lithospheric studies, with application to the West African craton beneath Nigeria. Three-dimensional MT inversion is the most powerful tool for imaging electrical conductivity structure from the surface to depths of hundreds of kilometres, but conventional smooth inversion produces blurred models that inadequately capture sharp geological boundaries such as fault zones, basin margins, and Moho discontinuities. This study proposes a novel sparse model parameterisation using a union-of-subspaces representation that combines smooth background models with spatially localised sharp features, implemented in a total variation regularisation framework with model basis adaptivity. The methodology is validated on synthetic models representing Nigerian tectonic scenarios including sedimentary basin margins and lower crustal conductors. Field application uses a 35-station broadband MT dataset acquired along a 250-kilometre profile across the West African craton-Benue Trough margin in Kogi and Kwara States. Inversions are performed on the IRIS HPC cluster using parallel finite-difference forward modelling. Findings demonstrate that the novel parameterisation recovers sedimentary basin margins with 40 percent sharper edge definition than conventional smooth inversion at equivalent data misfit. The lithospheric model reveals a major Proterozoic suture zone beneath the Benue Trough onset location characterised by a conductive mid-crustal body (less than 50 ohm-metres) at 15 to 25 kilometres depth, interpreted as a reactivated cratonic mobile belt boundary with graphitic metasediments. The study makes an original contribution to 3D MT inversion theory and its geological application in West Africa.

Keywords: magnetotelluric inversion, 3D lithospheric imaging, total variation, West African craton, novel parameterisation.