音频大地电磁法“死频带”畸变数据的Rhoplus校正

为研究AMT数据的"死频带"畸变特征及其校正方法,在长江中下游地区,利用凤凰MTU-5A仪器,系统观测了音频电磁场的"死频带"数据.对不同季节、不同时段的AMT"死频带"数据进行了分析.提出AMT"死频带"频率域畸变数据的Rhoplus校正方法,给出了该方法的适用条件、关键技术与评价方案,并提供了大量实测数据论证其应用效果.观测实验表明:在"死频带"内天然音频电磁场信号强度极低,正交电磁场分量相干度极低;造成了阻抗视电阻率、相位及相位张量等数据的畸变;畸变范围可达10k~100Hz,以5~1kHz频段最为明显;并且数据畸变程度与观测时段密切相关,在长江中下游地区为秋冬季畸变强夏季弱,日间畸变强夜间弱.处理实例表明:Rhoplus校正方法可以有效地处理AMT"死频带"内畸变的视电阻率、相位数据,相干度阈值与人机交互相结合的处理策略快速客观,所得结果曲线光滑连续,数据与实测未畸变数据基本一致.

Application of the Rhoplus method to audio magnetotelluric dead band distortion data

In the range 1- to 5 kHz, so called the audio magnetotelluric (AMT) ‘dead band’, signal power of the natural electromagnetic field has low values. It leads to the distortion of the impedance data and may cause misinterpretation. It is difficult to correct the distortion data with time series or power spectrum data processing methods. With a large number of experimental data, the Rhoplus method can be used to obtain a reasonable result.A series of AMT observation experiments were accomplished using the Phoenix MTU-5A instruments. The AMT data was recorded in different seasons and different days, some data was recorded both in day and night. On the basis of these data, firstly the distortion characteristics of AMT ‘dead band’ data were analyzed by using signal power, signal coherence, apparent resistivity, phase and phase tensor. Secondly the Rhoplus technique is applied to correct the distortion data. Its applicable conditions, processing scheme and key detailed techniques are presented. Finally, a large amount of data, including six sites and a test line, are provided to demonstrate the effectiveness of the method.The experiments show that the natural electromagnetic signal strength is rather low and the coherences of the orthogonal electromagnetic field components are low in the AMT ‘dead band’, which lead to the distortion of the impedance apparent resistivity, phase and phase tensor data. The frequency range of AMT ‘dead band’ distortion can be wide upto 10 k~100 Hz of which 5 k~1 kHz is most obvious generally. The effect of the distortion is closely related to the observation time and more serious in autumn and winter than summer, and more serious in the daytime than nighttime. The processing results show that when the main noise in the AMT ‘dead band’ is incoherent noise, the distorted data can be picked out automatically by using the coherence threshold value, and the Rhoplus method can obtain a good result that is consistent with the undistorted reference data which is observed in the night time or summer season. When the main noise in the AMT ‘dead band’ is coherent noise, the distorted data can be picked out manually by the interactive processing, the Rhoplus method can obtain a good result that is consistent with the undistorted reference data too. For the case of distorted data has no manifestation distinction with the undistorted data, the Rhoplus method can identify the distortion and correct it by using the high quality data in the frequency band over or under the AMT ‘dead band’. For the case of distorted data has a wide distortion band and poor quality, the Rhoplus method can be effective too by using the limited high quality data at some frequencies.AMT impedance apparent resistivity, phase and phase tensor data in the so called AMT ‘dead band’ may distort when observations are conducted in the daytime of autumn or winter season. The distorted data can be corrected effectively by the Rhoplus method, the combined data selection strategy of the coherence threshold value and interactive processing is fast and objective, the result curves are smooth and corrected data can be basically in line with undistorted observation data.