Deep images of electrical conductivity in Parnaiba basin - NE Brazil

Long period magnetotelluric (LMT) stations were deployed on an array format covering much of the Parnaíba basin and western edge of the Borborema Province in NE Brazil. A grid with 56 LMT stations (from 10 s to over 10⁴ s) with 70 km spacing were acquired during periods of acquisition varying from 2-3 weeks up to 6 months through two field campaigns between November 2018 and July 2019. This study is the first of this kind undertaken in Brazil, much in line with the American EarthScope, the Chinese SiinoProbe and AUSLAMP - Australia array initiatives suggesting the way forward for a comprehensive understanding of large 3D electrical structures of the continental crust and the lithospheric mantle in the entire country.  The results already published show that the resolution of the models obtained is comparable to the inversions of seismic tomography, the sensitivity of the MT method being superior in sensing the melting fraction, temperature and water content in the mantle. As an example, the 70 km spacing between EarthScope stations proved adequate to delineate the main structural features of the middle crust to the upper mantle of the United States. The Parnaíba Basin is a cratonic basin that has been formed by sedimentary mega sequences deposited along the Phanerozoic, after the formation of the Gondwana Supercontinent, with sedimentation expanding over the Borborema Province. The basin's evolution was conditioned by subsidence processes along unstable crustal areas at the end of the Brazilian cycle and demarcated by grabens, fault zones and magmatism. The shallow sediments of the basin (around 2.5 to 3.4 km maximum) however widely spread 60,000 km² cover prominent crustal features including the location of the trans-Brazilian lineament, limits of the Borborema Province, Amazon craton and other important shear and suture zones that may be important to better understand the evolutionary geodynamics of the supercontinents. The time series of electric and magnetic fields recorded in the field have been passed through extensive quality control analysis and then were robustly processed in the frequency domain generating good quality impedance tensor and tipper transfer functions. Currently we are concentrating our efforts on testing several parameters (e.g. mesh design, starting resistivity, damping and covariance analysis) involved in the 3D inversion nonlinear conjugate gradient ModEM code. Following the best practical procedures suggested by previous MT studies to avoid bad impact on the inversion results, misfits have been achieved nRMS ~ 2.2 in joint inversion of impedance tensor and tipper and nRMS < 2 for separate inversion of these transfer functions. Notwithstanding, preliminary resistivity models present good agreement with previous geophysical studies in the area and portray remarkable large-scale middle crust and deep conductive structures inside the covered area. Since there are several geodynamic processes associated with this area and electrical signatures may still exist the MT data set can be an important key to understand the evolution of the pan-African cycle and unravel new findings related to the formation of the Parnaiba basin.