高抗噪性三维体曲率分析技术及其在高陡地层发育区的应用

A 3Dcurvature attribute analysis method with excellent anti-noise property suitable for high steep formation

曲率属性可以较好地描述岩层的弯曲、褶皱、断层和裂缝等信息,但是由于曲率属性提取过程中需要计算地震数据的二阶微分,因此对噪声比较敏感,影响了结果的稳定性和可信度。在模型试算的基础上,提出了一种新的高抗噪性的体曲率计算方法。首先利用小波变换方法提取三维地震资料的相位余弦体;然后基于复地震道理论计算三维体视倾角属性,对视倾角进行中值滤波以压制微分计算中生成的噪声;最后利用两个方向的视倾角属性计算得到各类高信噪比的体曲率属性。模型实验和实际资料处理结果表明,该方法计算的曲率属性对噪声不敏感,适用于低信噪比资料的处理,特别是在陡倾角地层发育的地区,计算结果比C3等相干属性假象更少,能够真实地反映地下断层及岩性边界等地质信息。

Curvature attributes is useful in delineating faults,fractures,flexures and folds.However,in order to get curvature attributes,it is necessary to calculate the second-order derivative of seismic data.The derivative calculation process is sensitive to noise and the results are unstable and unreliable.In this paper we present a new 3D curvature attribute analysis method with excellent anti-noise property.Firstly,wavelet transform is used to extract phase-cosine attribute from seismic data.Then apparent dips are calculated based on complex seismic trace theory and filtered using media filter.Finally,many types of curvature attributes with high S/N data are got by combining apparent dips in both directions.Applications of this method to both synthetic and real seismic data show that this method has two advantages.First,it is insensitive to random noise and adapt to low S/N data.Second,it can effectively suppress the pseudo boundary information caused by discontinuity between the tilted strata,so this method is more suitable for steep dipping strata compared with C3 method.