Spike Detection and Processing of TDEM Signals using Wavelet Transform

Abstract

This paper presents a new method for spike detection, parameterization and processing in ground based Time Domain Electromagnetic (TDEM) signals. Time Domain Electromagnetic methods are used for mineral exploration by sub-surface conductivity measurement of Earth. The Earth’s response measured using inductive sensors are contaminated by spikes in form of sferics and due to load and phase shift in power distribution lines. The spikes present in the measured signal can significantly affect the depth versus conductivity profile of Earth. Time Frequency domain based spike detection can give spike localization in time. This paper gives a simple but powerful algorithm based on continuous wavelet transforms for quantifying spikes occurring in TDEM signals. Discrete wavelet transforms can be used for spike processing such that only the spiky portion of the signal can be treated without affecting remaining signal. The algorithm is applied to synthetic signals followed by application to field signals. Superiority of present technique over non-linear filter like median filter is brought out for both synthetic and field signals.

Keywords: Sferic; Spike factor; Time-Frequency; TDEM; wavelet.