Selection and application of wavelet transform in high-frequency sequence stratigraphy analysis of coarse-grained sediment in rift basin  

作　　者：Ling Li Zhi-Zhang Wang Shun-De Yin Wei-Fang Wang Zhi-Chao Yu Wen-Tian Fan Zhi-Heng Zhang 

机构地区：[1]National Key Laboratory of Petroleum Resources and Prospecting,China University of Petroleum(Beijing),Beijing,102249,China [2]College of Geosciences,China University of Petroleum(Beijing),Beijing,102249,China [3]Civil and Environmental Engineering Department,University of Waterloo,Waterloo,ON,N2L5Z5,Canada [4]SINOPEC Petroleum Exploration and Production Research Institute,Beijing,101599,China

出　　处：《Petroleum Science》2024年第5期3016-3028,共13页石油科学（英文版）

摘　　要：Wavelet transformation is a widely used method in high-frequency sequence stratigraphic analysis.However, the application is problematic since different wavelets always return the same sequence analysis results. To address this issue, we applied five commonly used wavelets to theoretical sequence models to document some application criteria. Five gradual scale-change sequence models were simplified from the glutenite succession deposition by gravity flows to form the fining-upwards cycle sequences(FUCS) and coarsening-upwards cycle sequences(CUCS). After conducting theoretical sequence model tests, the optimal wavelet(sym4) was selected and successfully used with actual data to identify the sequence boundaries. We also proposed a new method to optimize the scale of continuous wavelet transformation(CWT) for sequence boundary determination. We found that the balloon-like marks in scalograms of db4, sym4, and coif4 wavelet determine, respectively, the fourth-order sequence boundary, the thick succession sequence boundaries in FUCS, and the thick succession sequence in FUCS and CUCS. Comparing the sequence identification results shows that the asymmetric wavelets had an advantage in high-frequency sequence boundary determination and sedimentary cycle discrimination through the amplitude trend of the coefficient, in which the sym4 wavelet is the most effective. In conclusion, the asymmetry of wavelets is the first selection principle, of which asymmetric wavelets are more sensitive to sediment deposition by flood flows. The match of the wavelet between the sequence is the second selection principle, in which the correlation of time-frequency impacts the accuracy of sequence surface localization. However, the waveform of the wavelet is a visual and abstract parameter for sequence boundary detection. The appropriate wavelet for lacustrine sequence analysis is the asymmetric wavelet with a weak number of side lobes. The depositional flows, depositional process,and autogenic are three sedimentary factors that influence t

关 键 词：Wavelet analysis High-resolution sequence Sedimentary cyclicity Asymmetric wavelets 

分 类 号：TE311[石油与天然气工程—油气田开发工程]