矿集区大地电磁噪声处理方法及其应用

Noise processing methods and application study of MT in the ore concentration area

通过研究平移、线性插值拟合及基于经验模态分解(EMD)等去噪方法的原理,编写相应的人机联作与EMD分解去噪程序来处理实测大地电磁测深数据。发现:传统远参考方法在去除强能量干扰时能力有限,单纯的应用人机联作去噪也存在一定的不足,而在人机联作去噪的基础上再进行EMD分解去噪可以有效地压制噪声,从而提高矿集区大地电磁数据的信噪比。对比去噪处理前后的实测剖面反演结果,经过去噪处理的反演结果局部虚假异常得到有效的去除,为地质解释提供了更为准确的地球物理依据。

In this paper, the authors studied the de?noising method theories based on manual translation, linear interpolation fitting and EMD decomposition, compiled corresponding interactive and EMD decomposition de?noising programs, and then processed the measured MT data. Examples given in this paper show that the capacity of traditional remote reference method for removing the strong energy interference is limited, and simple application of interactive de?noising also has certain insufficiency. In contrast, EMD based on interactive de?noising can remove noise effectually and thus improve signal?noise ratio in the ore concentration area. Based on comparing the measured profile inversion results before and after de?noising processing, the authors point out that the false anomaly after de?noi?sing processing can be effectively removed, and hence more accurate geological interpretation basis can be provided.