基于蒙古东戈壁AMT数据的“死频带”分析

Dead band analysis based on Mongolian East Gobi AMT data

音频大地电磁测深是利用1 Hz^10 kHz频率范围内天然电磁场进行勘探的一种被动源电磁法.天然电磁场本身信号微弱,尤其是1 k^5 kHz有时甚至会低于仪器噪声水平,被称为"死频带"."死频带"电磁场信号微弱,且具有昼夜变化的特点,晚上信号强度强于白天."死频带"范围内正交的电场和磁场的相关度较低."死频带"的影响在视电阻率曲线上表现为在1 k^5 kHz范围内的频点出现严重脱节现象,但不同频点受到的影响程度不同,其中1500 Hz受影响最严重.文章利用在蒙古国东戈壁省安静电磁环境中取得的AMT数据充分分析了"死频带"的频谱和相关性等数据特征,及其对视电阻率和相位曲线的影响,并使用时间域去噪法处理"死频带"数据收到了一定的效果,即,选择满足约束条件的时间序列,删除不满足约束条件的时间序列,这里的约束条件可以是电磁场强度或者电场与磁场的相关度.具体做法为在时频变换的基础上,通过设置一定的临界值检索信号强度较大的特定频率的高频脉冲,然后重新组合成新的时间序列.最后对仪器的采集方案提出了改善意见.

Audio Magnetotelluric Sounding(AMT) is a passive source electromagnetic method for exploration using natural electromagnetic fields in the frequency range of 1 Hz^10 kHz. The signal of the natural electromagnetic field itself is weak, especially 1 k^5 kHz, sometimes even lower than the noise level of the instrument, which is so called "dead band". The dead-band electromagnetic field signal is weak and has the characteristics of diurnal variation. The signal intensity at night is stronger than that at daytime. The correlation between the orthogonal electric field and the magnetic field in the dead band is low. The effect of dead band on apparent resistivity curve shows that the frequency points in the range of 1 kHz to 5 kHz are severely disjointed. But, different frequency points are affected to different degrees, of which 1500 Hz is the most seriously affected. The paper fully analyzes the data characteristics of the spectrum and correlation of the dead band by using the AMT data obtained in the quiet electromagnetic environment in the East Gobi Province of Mongolia. Based on the understanding of these data characteristics, this paper uses time domain denoising method to process dead-band data and achieves good results. That is to say, the time series satisfying the constraints are selected and the time series not satisfying the constraints are deleted, where the constraints can be the intensity of electromagnetic field or the correlation between electric field and magnetic field. The specific method is to retrieve the high frequency pulses with high signal intensity by setting a certain threshold on the basis of time-frequency transformation, and then recombine them into a new time series. Finally, some suggestions for improving the acquisition scheme of the instrument are put forward.