三维稀疏反演多次波预测及曲波域匹配相减技术

传统的自由表面多次波压制(SRME)方法研究是基于二维假设的,这种算法对多次波贡献考虑过于局限,难以应对复杂地质构造的情况。3DSRME算法过于笨拙,对数据要求过于苛刻,很难有效推广和应用。三维稀疏反演多次波预测方法利用柯西概率密度函数,将垂直测线方向上稀疏的多次波能量贡献道集转换到模型空间,有效重构同相轴能量顶点——菲涅尔带,从而有效拾取多次波能量,准确预测多次波。采用曲波域匹配相减技术对多次波进行去除,相较于传统的最小平方匹配相减方法,Curvelet域具有多方向、多尺度的特性,能够更有效地压制多次波。建立三维倾斜层状速度模型,对模拟三维地震数据进行试算,并与常规2D SRME的处理结果进行对比,验证了本文所述方法的有效性。

Multiple Prediction with 3D Sparse Inversion and Curvelet Match

The traditional study of surface-related multiple elimination (SRME) is based on two-dimensional algorithm to confine the contribution of multiple waves,which is too constrained to deal with a complex geological structure.The 3D SRME algorithm can hardly be promoted and applied because of its clumsy algorithm and harsh demand of data.The multiple prediction with 3D sparse inversion takes advantage of Cauchy probability density function totransform the multiple waves' energy contribution gathers to model spacetoreconstruct the event energy vertex-Fresnel zone so as to calculate and predict multiple waves' energy effectively and accurately.The Curvelet match method is used to remove multiples from seismic data withretainingmany more directions and scales more effectively compared to the traditional least square matching phase subtraction.In comparison withthe result of 2D SRME,the effectiveness of this method is verified through establishing a 3D inclined layer velocity model and testingit on simulated 3d seismic data.