Signal Preprocessing And Induced Polarization Effect Recognition On Transient Electromagnetic Measurement

Transient electromagnetic (TEM) method establishes a pulse magnetic field underground through the ground electrode or the ungrounded loop with pulse current, and within the first pulse magnetic field intermittent, it observes the second eddy current electromagnetic field in the loop or the ground electrode in the time domain. This method has the advantages of deeper detecting depth, simpler observation and higher efficiency, so good effects are gotten in geological mapping, hydrocarbon, geothermal, metal mine, groundwater and permafrost exploration etc.Two kinds of signals existing in the transient electromagnetic measurement have been researched in this article. On the one hand, the gathered signals are often accompanied by lots of serious random disturbance noise (which have dynamically and transient characteristic), for the transient electromagnetic is the broad spectrum measurement. Moreover, the second fields signals are weak, which often submerge in the noise. Therefore, the research of signal preprocessing on the transient electromagnetism is the data interpretation foundation. On the other hand, To induced polarization effect that sometimes raise in middle or later measurement when there exist polarized earth or polarized object, the exploration research was carried out that could identified induced polarization effect, and used two kinds of effects EM and IP simultaneously. In this paper, the predecessor research results have been firstly reviewed, especially the induced polarization effect in the magnetic source TEM was carried on a comprehensive summary. Considered that there have been many research about magnetic source TEM, and the similarly forming mechanism of the magnetic source TEM and the electrical source TEM fields in some aspects, this article takes the signal preprocessing and the identifying induced polarization effect in the electrical source TEM as the main question.1. To signal preprocessing discuss, we focus on the two prime issues: (1) signal denoising, (2) prestack and stacked record analysis. In signal denoising, we choose five kinds of threshold (fixed threshold, stein's unbiased estimate of risk, heuristic threshold, and minimax threshold) to wavelet denoising in TEM signal. The comparison of their results showed that different kinds of threshold produce different effect on signal denoising, and adopt appropriate threshold can outstandingly increase signal-noise- ratio (SNR) that recover raw signal from interference noise. In prestack and stacked record analysis, we used wavelet analysis to multiscale analyze signals collected from China National Petroleum Company (CNPC), which observe the distribution characteristic of signal at different frequencies. The result shows that it is significant to make full use of the wavelet approximation of prestack record for the data processing and explanation in deep exploration.2. To identifying induced polarization effect discusses, we employed frequency dependant complex conductivityσ(ω) to describe polarizable earth with induced polarization effect in accordance with Cole-Cole model, applied time frequency analysis to transform 1D TEM signal into 2D time-frequency distribution, and then establish a identifying induced polarization signal method in TEM measurement, which classified by three kinds of quantitative indicators (Euclidean distance, absolute distance, and similarity) and six typical geoelectric section model of layered polarization. This method provides a new idea for eliminating IP effect on interpreting the electromagnetic response, and using of electromagnetic (EM) and induced polarization (IP) effects simultaneously.