The removal of the high-frequency motion-induced noise in helicopter-borne transient electromagnetic data based on wavelet neural network

Abstract

In helicopter-borne transient electromagnetic (HTEM) signal processing, removal of motion-induced noise is one of the most important steps. A special type of short-term noise, which could be classified as high-frequency motion-induced noise (HFM noise) based on its cause and time-frequency features, was observed in the field data of the Chinese Academy of Sciences-HTEM system. Because the HFM noise is an in-band noise for the HTEM response, it usually remains after the normal denoising procedure developed for the conventional motion-induced noise. To solve this problem, we have developed a three-stage workflow to remove the HFM noise using the wavelet neural network (WNN). In the first stage, the WNN training is performed, and the data segment in which the HFM noise is dominant is selected as the sample set. In the second stage, the HFM noise corresponding to the data segment in which the earth’s response coexisted with the HFM noise is predicted using the well-trained WNN. In the last stage, the predicted HFM noise is removed from the corresponding original data. As an example, we applied our workflow in the field data observed in Inner Mongolia, the HFM noise is removed effectively, and the results provide a strong data foundation for the subsequent processing procedures.

The Progress of the Helicopter-Borne Transient Electromagnetic Method and Technology in China

“…The high-frequency MIN is therefore a serious in-band noise in HTEM data. Wu et al [35] proposed a processing flow based on the wavelet neural network, which use the wavelet neural network to realize the effective modeling and prediction of high frequency MIN, so as to realize the accurate removal of it. In addition, Zhu et al [36] demonstrated that in addition to the main source of MIN, the movement of the normal vector of the coil in the secondary field also leads to MIN in the region with a strong secondary field.…”

Text Analysis Reveals Major Trends in Exploration Geophysics

“…A lot of research has been done in recent years to find the uses of neural networks in science in general and in geophysics in particular. There are many studies on the tasks neural networks can solve: reducing the noisiness of aerial electromagnetic surveys [21], automated fault prediction [22], prediction of laboratory earthquakes using machine learning [23], and many others. Neural networks in geophysics are used for a very wide range of tasks and this is really an unusual phenomenon.…”

The Development and Applications of the Semi-Airborne Electromagnetic System in China

“…In order to eliminate the influence of motion noise in SAEM observation, Liu et al [25] introduced the Ensemble Empirical Mode Decomposition (EEMD) method to decompose the data into several Intrinsic Mode Functions (IMFs), and the motion noise will be separated from the useful signal at different levels of IMF. In another related study, Wu et al [46] proposed a method based on the wavelet neural network in order to overcome the influence of high-frequency motion noise. This method was originally proposed for helicopter-borne TEM data, however, since SAEM observation data sometimes also contain the same phenomena, and the method has now also been introduced into the processing of SAEM data.…”