论文部分内容阅读
处理垂直地震剖面(VSP)的基本步骤之一,是区分上行波与下行波。考虑到这一点,我们利用这些波来模拟在钻井中记录到的地震资料,并且用数学方法导出了分离它们的最佳“最小二乘方”法。这个方法在实际上对于记录层次之间的间距没有什么限制,并且几乎能够完全消除相干下行信号。一维模型的精确公式要求将声阻抗信息包括在内,然而却有一种合理而现实的近似方法允许省略它们。我们推导了层的间距为规则的和随机的这两种有限条件下的频率一波数响应,并且同“常用”的速度滤波器的响应进行了对比。通过仔细地研究有效的测井资料,可以选择记录层以便能进行最佳对比;与此同时,在处理资料的编辑过程中,要求事先进行归一化和“真振幅恢复”,“真振幅恢复”校正可以更合理地根据声测井资料进行计算,这再次强调了在地震和测井方法之间实现更加完美的通力合作的可能性。
One of the basic steps in handling vertical seismic profiles (VSPs) is to distinguish between upstream and downstream waves. With this in mind, we use these waves to simulate the seismic data recorded in drilling wells and mathematically derive the best “least squares” method of separating them. This method has virtually no limit on the spacing between recording levels and almost completely eliminates coherent downstream signals. The precise formula for a one-dimensional model requires that acoustic impedance information be included, however there is a reasonable and realistic approximation that allows omission of them. We derive frequency-wavenumber responses for both finite and regular finite-layer conditions, and compare them with the response of a “common” velocity filter. By carefully studying effective well logs, the recording layers can be selected for best comparisons; at the same time, prior to normalization and “true amplitude recovery”, “true amplitude recovery ”Corrections can be more reasonably calculated from acoustic logging data, again emphasizing the possibility of achieving a more perfect co-operation between seismic and logging methods.