三维叠前相干干扰的正交多项式拟合压制方法

与二维叠前地震资料中的直线相干干扰不同，三维叠前地震资料中相干干扰通常为双曲线。常规相干干扰压制方法通常假设相干干扰为直线，因此在三维地震资料处理中就不再适用。本文利用正交多项式很好地拟合了直线型及双曲型相干干扰信号。在此相干干扰信号时空方向上通过加权叠加平均的方法压制相干干扰。由于此法没有对信号的频谱做任何切除，所以保留了所有的频谱信息。理论数据及实际资料处理结果表明：此法对二维叠前资料中的直线型及三维叠前资料中的双曲型相干干扰都有很好的适应性，相干干扰压制效果明显。

3D pre-stack coherent noise suppression with orthogonal polynomial fitting

Coherent noise on 3D pre-stack seismic data is generally hyperbolic, which is different from linear coherent noise on 2D pre-stack seismic data. So 3D coherent noise cannot be denoised by conventional 2D coherent noise denoiseing method. Here we propose an approach for 3D pre-stack coherent noise suppression with orthogonal polynomial fitting. We first fit linear and hyperbolic coherent noise with bivariate orthogonal polynomial fitting. Then we use the weighted average method to suppress coherent noise in their temporal and special directions. With no spectrum excision, this approach preserves all of the spectral information. Tests on model and real data show that this method can well remove both linear coherent noise on 2D pre-stack data and hyperbolic coherent noise on 3D pre-stack data.