Time-frequency analysis of seismic data using synchrosqueezing three parameter wavelet transform

High-quality time-frequency representation (TFR) is important for signal analysis. Due to the Heisenberg uncertainty principle, traditional time-frequency methods always lead to ambiguous TFR which has negative effect on signal analysis. In this paper, we introduce a sparse and invertible time-frequency analysis method called the synchrosqueezing three parameter wavelet transform (SSTPWT) using the three parameter wavelet (TPW), which is developed to improve the quality and readability of TFR by condensing it along the frequency axis. Numerical experiments on synthetic signals and real seismic data show its effectiveness.