next up previous [pdf]

Next: Acknowledgments Up: Chen et al.: Time-frequency Previous: Field examples


We have proposed a new TF analysis approach for seismic data using the synchrosqueezing wavelet transform. The proposed approach can achieve higher resolution both in time and frequency compared with the S transform and the local-attributes based TF decomposition. Numerical examples show that the SSWT can capture the variable-frequency components embedded in the non-stationary signal exactly. We show the potential of the SSWT in detecting anomalies of high-frequency attenuation, which is crucial for finding oil & gas traps. Field data tests also show that the proposed TF analysis approach can also help to detect deep-layer weak signal that is usually smeared in seismic data, which in turns can help to find deserted residual oil & gas.