Noise suppression of GPR data using Variational Mode Decomposition

Ground penetrating radar (GPR) has been used in the many aspects, such as civil engineering and the earth sciences. And the analysis and noise suppression of GPR data have always been the research focus. In this study, a new self-adaptive time-frequency decomposition tool called the variational mode decomposition (VMD) is introduced. We use the VMD method to derive a set of stationary sub-components, and based on the decomposition, we separate the valid signals and the components which are corresponded to the noise. One trace of GPR data are given to test the effect of the VMD decomposition, and the empirical mode decomposition (EMD) is also employed as a comparison. And a primary noise-suppression method based on the VMD scheme is also proposed. The application of the field GPR data further demonstrates the better performance of the proposed method in both noise suppression and the retention of geophysical events.