Interval Approach to Magnetotelluric Data Processing

Abstract

Abstract The article presents a new approach to estimate the frequency characteristics of the impedance tensor for processing magnetotelluric data. The approach is based on the applying of interval analysis methods when solving a system of linear equations. As a reference method, to compare with, a combined robust algorithm is used (with discarding data by the coherence criterion, median estimating, and weighting least squares method). This algorithm is compared with the results of the proposed interval computational algorithm that is based on the method of J. Rohn, implemented in the intvalpy Python library. Computational experiments on the data processing were performed using natural magnetotelluric field data. The interval approach can be successfully applied to the processing of magnetotelluric data.