Use of principal component analysis in the de-noising and signal- separation of transient electromagnetic data

We present a method based on principal component analysis (PCA) for suppressing noise and separating signal sources in transient electromagnetic (TEM) data acquired in environmental clean-up and hydrogeophysical applications. Coherent noise due to background responses and incoherent noise often contaminate data in these applications and preprocessing is required for quantitative analyses. In these problems, either central-loop configuration over closely spaced stations or fixed-loop setup with multiple stations is used. Resulting data are ideal for decomposition by PCA. In this paper, we outline the basics of PCA and apply it to the processing of TEM data from unexploded ordnance (UXO) clearance. We demonstrate that PCA can clearly separate the background geologic noise due to magnetic soil and suppress incoherent noise contaminating data in late time gates.