Transient Re-Extracting Transform and its application in seismic thin layer characterization

In oil and gas exploration, thin-layer reflection signals are prone to tuning effects, compromising the accurate extraction of amplitude and frequency and resulting in misjudgment of stratigraphic relationships. Transient Extracting Transform (TET) is an effective spectrum analysis tool that significantly improves the transient characteristics of seismic signals by rearranging and extracting time-frequency spectrum coefficients. However, TET suffers from chaotic rearrangement point distribution in thin-layer signal characterization, and is seriously affected by inter-layer interference in seismic data analysis. Therefore, this paper proposes an improved Transient Re-Extracting Transform (TRET). This method analyzes the extraction process of TET in the short-time Fourier transform spectrum, combines the convergence of rearranged estimation points, and derives the fixed-point conditions on the group delay curve. Based on this, a differential convergence constraint Re-Extraction Operator (REO) is constructed to eliminate the local divergent group delay points in TET, effectively suppressing the confusion of interface reflected signals and distortion of interpretation results when processing TET in thin-layer signals due to the small spacing between group delay traces. The effectiveness of TRET is verified through two synthetic models and one field seismic data, indicating that this method can provide clearer and more stable characterization results along the time axis, enhance interface recognition capabilities, and provide a reliable indication for thin layers and their interface analysis and characterization, which has important practical value for the exploration and development of complex oil and gas reservoirs.