多尺度迭代地震高分辨率反演算法在W地区薄储层识别中的应用

Application of Multiscale Iterative Seismic High-Resolution Inversion Algorithm in Thin Reservoir Identification in the W Region

摘要: 陆相页岩气薄互层发育、非均质性极强、岩性组合复杂，目前现有的地震反演分辨率无法满足精度需求。本文利用基于频谱延拓的高分辨率处理技术，构建频变频谱拓展因子及高分辨率处理传导算子，将其应用到地震数据处理中，数据频带宽度由8~70 Hz扩展到5~120 Hz，地震数据分辨率显著提升。在此基础上，通过小波变换多尺度分解技术，使得低、中、高频段的地震信息能够充分利用，有效缓解了常规反演方法高低频缺失的问题，高低频相融合提高反演分辨率，实现W地区陆相页岩气薄储层预测。

Abstract: The continental shale gas in the W region is characterized by thin interbeds, extreme heterogeneity, and complex lithologic combinations. The existing seismic inversion techniques cannot meet the accuracy requirements due to the poor resolution. In this paper, a high-resolution processing technique based on spectrum extrapolation is used to construct a frequency-dependent spectrum extension factor and high-resolution processing operator, which are applied to seismic data processing. The data frequency bandwidth is extended from 8~70 Hz to 5~120 Hz, significantly improving the seismic data resolution. Based on this, the multi-scale decomposition technique of wavelet transform is used to fully utilize the seismic information in the low, medium, and high frequency bands, effectively alleviating the problem of missing high and low frequencies in conventional inversion methods. The fusion of high and low frequencies improves the inversion resolution and achieves the prediction of thin reservoirs of continental shale gas in the W region.