LAUSR

Abstract Discrete wavelet transform (DWT) has been widely used as a useful tool in denoising geophysical data for its outstanding feature of detecting singularities and transients. In this paper, we develop a new strategy of denoising borehole transient electromagnetic (BHTEM) data. The principle idea of the denoising process is keeping those coefficients which are necessary to reconstruct the signal unchanged and setting others to zero. In our case, according to results of modeling, only the first eight detail coefficients are needed. The method has been validated on both synthetic and field BHTEM raw data. Besides the DWT, complementary measures are introduced accordingly. For simple data set, a curve fitting technique is employed to smooth the signal furthermore. For field BHTEM raw data the correlation analysis is carried out to identify and correct distorted transients. The efficiency and reliability of the method are proven by the results.