逆时偏移数据体低波数噪声相对保幅衰减技术及应用

Relative amplitude-preserving attenuation technology of low wavenumber noise in reverse time migration data volume and its application

逆时偏移成像过程中压制低波数噪声后的叠加剖面仍存在较强能量的低波数噪声,其掩盖了有效反射信号对地质目标细节的精细刻画,因此,相对保幅地恢复有效反射信号是实现逆时偏移技术实际应用的重要保障。为此,将热力学领域中三维扩散滤波方法引入到深度域逆时偏移低波数噪声压制中,并对含低波数噪声的三维逆时偏移成像数据体进行体迭代处理。通过优化迭代次数和扩散系数2个关键参数,实现了低波数噪声和有效反射信号的相对保幅分离,同时采用局部化数据体输入和临时输出存储的优化策略,突破了计算节点内存限制,构建了一种高效的、相对保幅的低波数噪声衰减技术。理论模型和实际地震数据应用结果均表明,该技术处理结果的保真保幅性明显优于国际上常用的拉普拉斯算子去噪方法,且对噪声的敏感性弱,可以在深度域数据体上有效恢复掩盖在低波数噪声能量之下的有效反射信号,实现复杂地质目标的地震精细刻画,且其处理效率能够满足实际生产周期的需求,这可为同类深度域含低波数噪声成像资料的地震精细刻画提供方法和技术参考。

During reverse time migration imaging,the energy of low wavenumber noise is still strong in the stacked section after suppression,which hinders the fine description of the details of geological targets by effective reflection signal.Therefore,how to recover the effective reflection signal with relative amplitude preservation is vital for the practical application of reverse time migration technology.The three-dimensional diffusion filtering method from the field of thermodynamics is introduced for the first time into the suppression of low wavenumber noise in reverse time migration in the depth domain,and volume iterative processing is performed on the three-dimensional inverse time migration imaging data volume containing low wavenumber noise.By optimizing the number of iterations and diffusion coefficient,the low wavenumber noise and effective reflected signal are relatively separated with amplitude preservation.Furthermore,the optimization strategy is adopted to break through the limitation of the computing node’s memory size by input of localized data volume and output of temporary data storage.Thus,an efficient low wavenumber noise attenuation technology is constructed.The theoretical model and practical data application show that the proposed technology is better than the Laplace operator method in terms of fidelity and amplitude preservation,and the sensitivity to noise is weaker.It can effectively recover the details of effective reflection signal from the depth domain data volume,which is covered under the low wavenumber noise energy,and the fine seismic characterization of complex geological targets.The processing efficiency can meet the needs of practical data processing cycles.Therefore,it can provide a technical reference for the fine characterization of seismic imaging data with low wavenumber noise in the depth domain.