Automatic extraction algorithm for whistle wave propagation parameters

Whistler waves serve as vital probes for investigating the space physical environment, with their propagation parameters being essential for understanding their wave characteristics. However, current manual extraction methods are inadequate for handling the scale of modern observational data. To address this issue, we proposes a new automatic extraction algorithm for whistler wave propagation parameters. First, a 4-second sliding time window is used to process the ELF band magnetic field data from the Zhangheng-1 satellite (ZH-1) to obtain waveform segment data A. Second, a short-time Fourier transform is applied to generate the time-frequency spectrograms B. Third, a target detection network is then employed to automatically identify the time-frequency range C of the whistler waves from the spectrograms. Fourth, C is used to isolate the corresponding whistler-containing waveform segments AA from A. Fifth, automated noise suppression based on C is performed on AA, yielding denoised waveforms AA′. Sixth, the high-precision magnetometer (HPM) to the time range of AA′, along with AA′ data, are then input into a wave vector analysis algorithm to obtain the propagation parameters of the whistler waves. Finally, the model is evaluated using data from the ZH-1 from June 2020. The results indicate that the detection accuracy of whistler waves reaches 95%, and the cosine similarity between the automatically extracted propagation parameters and the manually extracted results is 0.99, demonstrating that the accuracy is comparable to that of manual methods. This approach provides new technical support for the intelligent analysis of space electromagnetic wave propagation characteristics.