【摘要】 本文提出了基于波原子域的叠前地震资料信噪分离方法,首先在波原子域进行多尺度分解,然后根据能量集中原则,采用不同尺度间系数相关性分离有效信号和噪声。数值模拟及实际资料的处理表明,基于系数相关性阈值的波原子域叠前地震资料信噪分离方法可以有效压制随机噪声,同时能兼顾抑制部分相干噪声。处理后地震剖面的信噪比得到了大幅提高,反射同相轴更加连续,为后续的处理解释提供了高信噪比的数据。更多还原

【Abstract】 The signal-noise separation method for pre-stack seismic data in the wave atomic domain was proposed in this paper,at first the multi-scale decomposition was conducted in the wave atomic domain,then based on the energy concentration principle the coefficient correlation for the different scales were utilized to separate effective signal and the noise.The results of numerical simulation and field data processing indicate that the pre-stack seismic data signal-noise separation method which is based on the coefficient correlation threshold in the wave atomic domain could effectively suppresses the random noise,meanwhile it also could restrain some coherent noise.It can be seen that the signal to noise ratio for the processed seismic sections was greatly raised and the reflection events were more continuous,providing high signal to noise sections for the subsequent processing.更多还原