基于反射波各向异性特征的保真去噪方法

Seismic fidelity de-noising with reflection anisotropy

联合地震数据的L_2范数以及各向异性函数所表达的地震波组特征的L1范数,定义了空间域加权叠加系数,将传统的地震信号非均匀叠加转化为均匀叠加,压制随机噪声、改善同相轴连续性,提高地震资料后期处理解释的精度。针对地震资料保真去噪的要求,基于反射波的各向异性与随机噪声的各向同性特征,构造各向异性函数并通过加权叠加数据的保真实现地震数据的压噪。模型及实际资料的处理结果表明,该方法能够有效压制随机干扰,尤其对弱反射波的处理具有较好的保幅效果。

Combing the L2norm with the L1norm of seismic data, we propose a weighting function which transforms routine non-uniformity stacking for seismic data into uniformity stacking to suppress random noise and enhance the continuity of the reflection events. This approach improves seismic data precision for later data processing and interpretation. As reflection waves show anisotropic and noise is isotropic, we construct an anisotropic function and apply it in the weighted stacking to suppress noise based on the fidelity de-noising. Model and seismic data tests show that the proposed method can not only effectively suppress random noise but also preserve seismic amplitudes even weak reflection. © 2017, Editorial Department OIL GEOPHYSICAL PROSPECTING. All right reserved.