Abstract: It is difficult to accurately describe the thin reservoirs in unconformity traps with strong seismic reflections. Taking the Mesozoic unconformity trap in the Changdi area in the Jiyang depression as a case, the method for strong reflection separation is studied, the separation process is established, the selection of key parameters in the separation process is discussed, and the models are established to verify the feasibility of the separation method and select sensitive parameters for reservoir prediction. The study shows that the matching tracing algorithm with local frequency as a constraint can effectively avoid outliers, enhance the stability and increase the speed of calculation. Considering well logging data and near-well seismic data, it is proved that a 0.6 λ (scale factor) can find out strong reflections and highlight effective reservoir information in the target area. In the separated data volume, there is a positive correlation between the amplitude attribute and the effective reservoir thickness, which can reflect the lateral distribution characteristics of the reservoir. The drilling results of the wells are consistent with the predictions, which verifies the reliability of the method.

Key words: strong seismic reflection, matching tracing, local frequency, scale factor, effective reservoir, Changdi area