论文详情
地震数据包络的对数形式可以反映波形的宏观变化,同时,能有效提取地震数据的低频信息。对数包络所提取的低频信息与地震数据的主频具有相互对应关系。结合“分频多尺度”的反演思想,先对高频数据作低通滤波处理,再计算其对数包络以提取地震数据的低频成分,并以此构建目标函数进行反演,从而得到较好的初始速度模型。数值模拟反演实验证实:在全波形反演过程中,如果采用分频多尺度反演方法,可以有效缓解“周波跳跃”现象对反演结果精度的影响,准确反演出地下复杂结构的速度模型。
A logarithmic envelope of seismic data can reflect the macroscopic changes in seismic waveforms and can help obtain low-frequency signals; the extracted signal is positively correlated with the dominant frequency of seismic data.This paper proposes a full-waveform inversion method utilizing a multi-scale inversion strategy based on the logarithmic envelope of seismic data.First,low-pass filtering is performed on high-frequency data.Then,the logarithm envelope is calculated to extract low-frequency signals.Next,a logarithm envelope-based objective function is constructed to perform inversion for the initial velocity model.The numerical simulation inversion tests confirm that in the process of full waveform inversion,the proposed inversion method can effectively suppress the cycle-skipping phenomenon,improving the inversion accuracy of the velocity model for complex structures.
国家自然科学基金(41074089,41774135)、中央高校基本科研业务费专项资金(310826172002)共同资助。