Polarisation analysis of magnetotelluric time series using a wavelet-based scheme: A method for detection and characterisation of cultural noise sources

• Polarisation attributes in the time-frequency domain reveal EM signal features.
• A wavelet scheme is used to achieve the time-frequency characterisation.
• Synthetic signals, data from a field test and MT real data have been analysed.
• MT natural signal and cultural noise present different polarisation attributes.
• New technique applicable to the characterisation of controlled sources in general.

The identification and elimination of cultural noise that affects magnetotelluric (MT) time series presents a challenge in the vicinity of industrialised, urban or farming areas. Most noise sources are fixed in space and create a signal with certain polarisation properties. In this paper, we propose a new method for detection and characterisation of cultural noise sources in magnetotelluric time series based on polarisation analysis of the electromagnetic signal in the time–frequency domain using a wavelet scheme. We tested the proposed method with synthetic polarised signals and experimental time series corresponding to a field experiment with a controlled EM source and several MT real cases. The results demonstrated the difference between the polarisation properties of the natural MT signal and the signal contaminated by a controlled source or by cultural noise.