Ultralow-Frequency Geomagnetic Signal Estimation: An Interstation Transfer Function Method Based on Multivariate Wavelet Coherence

Electromagnetic perturbations associated with earthquakes and volcano eruptions have been intensively documented in the past decades, from both ground and satellite observations. However, the magnetic signals associated with crustal activity are usually very weak and mixed with global geomagnetic signals originating from external sources. Thus, one of the key issues in seismo-electromagnetic study is to identify local geomagnetic signals from global magnetic pulsations. The interstation transfer function can recover the global magnetic pulsations at the observatory by using the data of reference station, providing an effective way for signal discrimination. To further improve the accuracy of global magnetic signal estimation, in this study, we develop a new method for interstation transfer function calculation based on multivariate wavelet coherence using adaptive selection of time window lengths for different periods. Test on real data demonstrates that the global signals of the external source are accurately estimated, including the horizontal X and Y components and the vertical Z component of the observatory. The results of using different reference stations confirm the robustness of the proposed method. It is proved that the method can be used to eliminate the external ionospheric source signals effectively and identify local magnetic field signals lying in the background at the observatory. The proposed method can be useful in seismo-electromagnetic signal identification and extraction.