音频大地电磁法强干扰压制试验研究

The denoising of the audio magnetotelluric data set with strong interferences

压制非相关噪声已经有多种比较成熟的技术方法,如多次叠加、远参考、Robust阻抗估计等,但这些方法对压制矿集区存在的强电磁噪声(通常属于相关噪声)基本没有效果.本文在前期提出的针对矿集区大地电磁强噪声压制的数学形态滤波基础上,开展了音频大地电磁法强干扰压制的试验研究.在四川西昌某AMT测点附近,布置了大功率电磁发送站,通过接地导线向地下供20A的方波电流,同时在供电期和间歇期用凤凰公司MTU-5A仪器观测4个水平分量的电磁场.在室内采用数学形态滤波和阈值法相结合的手段在时间域对含有人工源强干扰的电磁场数据进行处理,以视电阻率对比、频谱和极化方向等参数对处理效果进行评价.结果表明:结构元素恰当的数学形态滤波可以对人工源强干扰进行识别、去除,阈值法则可以消除形态滤波后的脉冲干扰,二者结合可以有效地压制AMT中的强干扰噪声,提高数据可靠性.本文结果对AMT实际资料处理具有重要的实用和参考价值.

The interference is a challenging issue in the magnetotelluric（MT）method.There are several well-established techniques to suppress non correlative noises in the AMT methods,such as the multiple stacking method,the remote reference technique and the Robust impedance estimation algorithm.However,the above mentioned methods become inefficient for signals with strong interferences,especially for the AMT data set collected in the noisy ore concentratedareas.In this paper,we carry out an experiment of applying the mathematical morphology filtering for the noisy AMT data set.This field experiment is located a place nearby the city of Xichang in the Sichuan province.A square wave current of 20 Awas injected into the transmitter to generate the electromagnetic induction in the Earth.The Phoenix MTU-5Ainstrument is adopted to measure both horizontal electric and magnetic fields during the period with or without current on in the transmitter.We processed the noisy AMT data by the mathematical morphology filtering and threshold method and analyzed its effect on the apparent resistivity and phase,spectrum and polarization directions.Results show that mathematical morphology filtering with reasonable structural element can identify and suppress the strong interference.The threshold method can suppress the impulse interference after mathematical morphology processing.The combination of two methods can suppress the strong noise and improve the data reliability.