Joint inversion of resistivity and permittivity for two dimensional RMT data based on FCM clustering

Radio-magnetotelluric (RMT) is a new technique for shallow geologic survey. The frequency range of RMT is from 10 kHz to 1MHz, between the highest frequency of audio-magnetotelluric (AMT) and the lowest frequency of conventional ground penetrating radar (GPR). The effect of displacement current on RMT data increases with frequency and is related to electric permittivity. Therefore, the quality of RMT inversion model is significantly reduced with a high frequency when only considering conductivity in the inversion. A novel RMT inversion method which considers both conduction and displacement currents and recovers resistivity and permittivity properties simultaneously is proposed in this paper. To improve the mutual constraints between resistivity and permittivity, a fuzzy C-means (FCM) clustering constraint term is added to the classic minimum structure regularization function. The synthetic examples show the recovered resistivity and permittivity models are closer to the true model parameters after including fuzzy C-means clustering constraint term. The new method not only improves the effectiveness of RMT inversion but also recovers additional physical parameters which carry valuable subsurface geologic information.