Analysis of magnetometer data using wavelet transforms

Geomagnetic eld-line resonances may be identi ed via the use of cross-phase analy-sis of data from two closely spaced meridional ground-based magnetometer stations. It has been demonstrated that preprocessing of the data using a wavelet-based lter, chosen with regard to the variance of coe ¯ cients of the wavelet components, can be used to remove both low-frequency trend and large amplitude, localized struc-tures, thus facilitating the selection of the eigenfrequency. We demonstrate that the dependence of the variance of coe ¯ cients on the wavelet level during a geomagnetic storm event is characteristically di¬erent from that obtained from data on a day associated with quiet geomagnetic activity. This suggests that the spectral nature of magnetic disturbances excited in the magnetosphere is dependent upon the level of geomagnetic activity