利用自适应迭代多级中值滤波法分离海上多震源混合波场

Multi-source blended wavefield separation for marine seismic based on an adaptive iterative multi-level median filtering

多震源激发技术可提高海上地震勘探作业效率,但其混合波场的分离是影响成像的关键因素。伪分离记录中次震源波场在共炮集域之外的其他时域中呈现与主震源波场不同的随机分布特征。本文采用自适应迭代多级中值滤波去噪法以实现海上多震源混合波场的分离,迭代过程中通过自适应减小中值滤波窗口逐步提取波场细节信息。海上实际多震源采集混合炮集记录的波场分离结果表明,分离后的各震源单炮记录清晰,频谱特征相近,证实本文方法具有很好的实际应用价值。

Multi-source simultaneous acquisition improves greatly marine seismic operation efficiency.However the blended wavefield separation is a key factor for good imaging of this kind of data.When pseudo blend records are sorted from common shot gathers to other domains,aliasing of the secondary source is transformed into random distribution noise.In this paper, we propose multi-source wavefield separation based an adaptive iteration multi-level median filtering.Small window size of the median filter can extract more details of the wanted wavefield.Real blended data tests show that after blended wavefield separation by the proposed approach,single shot records of every source is clear and their frequency spectrums are similar,which proves the validity of the proposed approach.