摘要
In transversely isotropic media with a vertical symmetry axis （VTI）, the converted-wave （C-wave） moveout over intermediate-to-far offsets is determined by four parameters. These are the C-wave stacking velocity Vc2 , the vertical and effective velocity ratios γ0 and γeff, and the anisotropic parameter χeff. We refer to the four parameters as the C-wave stacking velocity model. The purpose of C-wave velocity analysis is to determine this stacking velocity model. The C-wave stacking velocity model Vc2, γ0, γeff, and χeff can be determined from P-and C-wave reflection moveout data. However, error propagation is a severe problem in C-wave reflection-moveout inversion. The current short-spread stacking velocity as deduced from hyperbolic moveout does not provide sufficient accuracy to yield meaningful inverted values for the anisotropic parameters. The non-hyperbolic moveout over intermediate-offsets （x/z from 1.0 to 1.5） is no longer negligible and can be quantified using a background γ. Non-hyperbolic analysis with a γ correction over the intermediate offsets can yield Vc2 with errors less than 1% for noise free data. The procedure is very robust, allowing initial guesses of γ with up to 20% errors. It is also applicable for vertically inhomogeneous anisotropic media. This improved accuracy makes it possible to estimate anisotropic parameters using 4C seismic data. Two practical work flows are presented for this purpose： the double-scanning flow and the single-scanning flow. Applications to synthetic and real data show that the two flows yield results with similar accuracy but the single-scanning flow is more efficient than the double-scanning flow.
在具有垂直对称轴横向各向同性介质中,利用四种参数来确定中间至远偏移距转换波(C-波)动校正。它们是C-波叠加速度VC2,垂直速度比和有效速度比γ和γeff,以及各向异性参数χeff。我们将这四种参数作为C波叠加速度模型。C-波速度分析的目的就是确定这种叠加速度模型。C-波叠加速度模型VC2,γ0,γeff,和χeff可以由P-波和C-波反射动校正资料获得。然而错误的传播是C-波反射动校正反演中的严重问题。当前短排列叠加速度由于是从双曲线动校正推算而得,因而其精度不足以为各向异性参数提供有意义的反演值。中间偏移非双曲线动校正不再被人们所勿略,而是可以用一个背景γ加以量化。非双曲线分析通过中间偏移距的γ校正量可以产生VC2,若数据不含燥音,其误差小于1%。方法稳健,允许γ启始假定值的误差达20%。该方法也适用垂直非均匀各向异性介质。精度的提高使能够用4分量地震资料计算各向异性参数。为此提出了两种工作流程:双扫描和单扫描流程。理论数据和实际数据的应用表明这两种流程得出的结果其精度相似,但是单扫描流程比双扫描更有效。
基金
This work is funded by the Edinburgh Anisotropy Project of the British Geological Survey.
