Improving magnetotelluric data-processing methods

Abstract

A magnetotelluric (MT) data processing algorithm that demonstrates high robustness to intense electromagnetic noise that occurs in measured MT data has been developed. The key features of the algorithm are a specific approach to estimating different transfer functions and the ability to utilize all four channels acquired at remote reference stations. The code utilizes various techniques to reduce the estimation errors, including the robust Huber estimator, jack-knife approach, improved remote reference technique, and compensating for overestimation of power spectra. The proposed algorithm has high efficiency in processing data with a low signal-to-noise ratio.