Abstract: The detection signal of the ground-airborne frequency-domain electromagnetic method (GAFDEM) is multi-frequency non-stationary signal, when extracting its amplitude using conventional Fourier transform methods, the resolution is poor. To solve the above problem, this paper proposes an amplitude extraction  method of ground-airborne  frequency-domain electromagnetic  signal  based on synchroextracting transform (SET). This method  performs SET on the electromagnetic data to obtain high-resolution time-frequency graph, and uses energy operator to make the energy of time-frequency spectral electromagnetic data more concentrated. At the same time, the greedy algorithm is used to realize the ridge extraction method to obtain the high energy band of the time-frequency graph. By filling in zero values in the ridges adaptively through autoregressive model, the endpoint effect problem caused by the window function is solved.  according to the complex value of the time-frequency-diagram at the ridge location, the variation of the amplitude of each frequency component with time is obtained, and the accuracy of the amplitude extraction of electromagnetic signals in the ground-airborne frequency-domain based on SET is investigated under different signal-to-noise ratios. Through calculation, when the signal-to-noise ratio  is greater than or equal to 10 dB, the average relative root mean square error  of the amplitude extraction result is less than 5%; When the signal-to-noise ratio  is less than 10 dB, the average relative root mean square error of the amplitude extraction result is less than 10%, achieving good extraction results. This method is applied to the GAFDEM delection  of the Huola Mountain tunnel project in Xinjiang, and the amplitude of each frequency component of the multi-frequency electromagnetic signal is successfully extracted. Compared with the method of extracting non-stationary signal amplitude using the Fourier transform, this method effectively improves the resolution of amplitude extraction results.

Key words: ground-airborne frequency-domain electromagnetic method, time-frequency analysis, nonstationary signal, parameter estimation, synchroextracting transform