The Study On Restraining Environmental Noise Of Magnetotelluric Sounding In Lengshui Ore Concentration Area

Electrical structure of Lengshui ore concentration area is studied using magnetotelluric (MT) method, understand the formation mechanism of the metallogenic system and forecast deep target. There are extensive noise pollutions in ore concentration area, which leads serious noise interference to the MT signal. How to reduce noise will cause MT surveys in ore district successful or failure. Noise characteristics, noise source in the ore concentration area and denoising technology of time‐domain MT signal is studied, data processing and inversion is done, and the electrical structure is established preliminarily in this paper.Noise characteristics in ore concentration area are studied based on measured audiomagnetotelluric method(AMT) data. Noise can be divided into4categories including pulse noise, periodic noise, square wave noise, triangular wave noise. The frequency range of noise is wide and the noise energy is generally strong, which almost completely destroy useful signal. Daily work of mining equipment, wireless communication, neighborhood stray current, power lines and traffic in the ore concentration area are the main noise source. De‐noising technology research include: To study the wavelet analysis method, the method of time‐frequency analysis of Hilbert‐Huang transform and matching pursuit method. The interference of some common interference in magnetotelluric signals was studied by using different SNR(Signal to Noise Ratio) models. Comparing with the advantages and disadvantages of above three methods, De‐noising technology has been found to deal with MT data disturbed seriously by the ore concentration area. Taking a set of measured magnetotelluric data as example, three groups of apparent resistivity and phase curves weres calculated by using measured magnetotelluric data that was handled without other de‐noising treatment and with wavelet analysis approach and method based on EMD threshold de‐noising and matching pursuit method de‐noising approach, results showed good de‐noising effect.In this paper, audiomagnetotelluric method data of139lines selected in the region was processed. Near to the ore concentration area, there was a variety of noise jamming. Data that was not interfered too large was eliminated with time series analysis by empirical mode decomposition method; but the measuring points disturbed very seriously were de‐noised with time series analysis by matching pursuit method. The profiles of magnetotelluric depth‐resistivity were inverted between de‐noising before and after, respectively. Results showed that although the inverted resistivity profile without de‐noising can reflect roughly hierarchical structure and low‐resistivity anomalous bodies, but not sufficient resolution to it. The inverted resistivity profile after de‐noising indicated that the electrical structure was clear, it can reflected not only the inserting intrusive rocks of high resistivity on the left, but also can draw the outline of low‐resistivity anomalous bodies more clearly.