Comparison of Fourier and wavelet techniques in the determination of geomagnetic field line resonances

Fourier transforms have traditionally been used in the past for time‐frequency analysis of numerous physical problems. Recently, a new time‐frequency analysis technique using temporally confined base functions, called wavelets, has been applied to many problems. The advantage of the wavelets over the conventional sinusoidal functions is their time localization property, providing information not only about the frequency or scale size of the features present but also about their location in the time series. Here we compare the performance of the two techniques on the time‐frequency analysis of ground magnetometer data and especially their ability to pick up the field line resonance (FLR) frequency of a resonating magnetic field line using automated FLR determination techniques. We find that the automated techniques work better with the Fourier transforms giving less variable FLR frequency. The high temporal resolution of the wavelet transforms can be useful for detailed analyses of short time periods, but it becomes a disadvantage when it comes to automated techniques for the entire dayside magnetosphere, yielding a highly variable FLR frequency.