The strong seismic reflection separation method based on the matching pursuit algorithm has yielded encouraging results in practical applications.However,this method faces some challenges in the selection of key parameters and has been scarcely investigated.This study thoroughly investigated the optimization of frequency parameters in the method thereof and established a technical process.First,this study analyzed the spatial variations in the frequency of seismic waves from the perspective of seismic wave propagation and attenuation and proposed a matching pursuit algorithm under the constraints of frequency.Then,this study designed corresponding geological models,conducted frequency optimization tests,and separated the strong seismic reflections using the matching pursuit algorithm under the constraints of the seismic wave frequency of the overlying strata.The test results verify the accuracy and effectiveness of the strong reflection separation method.Finally,the strong reflection separation method was applied to the fracture-vug carbonate reservoirs in the Tarim Basin.The great application effects indicate that the separation method can separate strong reflection events more reasonably and effectively,making the characteristics of underlying small-sized fractures and vugs more significant.This study further optimizes the strong reflection separation method based on matching pursuit and improves the adaptability and application performance of seismic data,thus laying a better foundation for further promotion and application of this method.