SEGJ Technical Conference

Robust impedance estimator based on Hilbert-Huang transform to time series magnetotelluric data

Abstract The magnetotelluric (MT) method using natural electromagnetic signals is useful to visualize the deep subsurface structure. The first step of MT data processing is transforming the time series data into the frequency domain data, and then calculating the impedance tensor. MT time series are known to occasionally contain non-stationary signal and temporal noise, in which case the time series are not suitable for the basic requirements of conventional methods based on the Fourier transform. In this paper, the method based on the Hilbert-Huang transform (HHT) is used for estimating the MT response functions from the time series of the electromagnetic field to minimize the estimation bias due to the non-stationary signal and temporal noise of MT data. To show the superiority of HHT compares with FFT in the MT impedance tensor estimation. At the step of impedance estimation, to keep the similarity with the conventional method (BIRRP, Chave et al., 2004), the robust iteratively reweighted least squares regression with the Huber weight function is used. Finally, field MT data affected by a geomagnetic storm are used to evaluate the effectiveness of the new technique, which was named RMHHT (Robust M-estimator based on HHT).