Abstract: Taking on denoising processing to the EM38 data from Mandamah Area in New South Wales, using wavelet transformation, then acquire the after-transformation values which have contained useful information and in which noises have been eliminated according to decomposing and restructuring. Compare the after-transformation EM38 values with the abnormal characteristics of Cu element, have an analysis and make the forecast. This method is effective in removing noises while retaining the useful information. The practical application result indicates that wavelet-analysis could play a special role in processing EM38 statistics and giving satisfying results.