论文部分内容阅读
我们提出了一种两步反演程序以便从大偏移地震资料中选取地下速度估算。第一步是自动迭代大偏移反演方法(ALLOIM),在海底有强多次波出现时,我们用这种方法做过试验。因为该方法是依赖于找到临界反射点的位置,所以就无法分辨低速带,而重新构成的仅仅是速度的平均估算。在存在低速带时就需要进行反演过程的第二步。这第二步就是用于亚临界反射的广义最小平方法。我们应用扰动技术来计算反演解,并从第一步确定的平均速度估算来确定参数模型。这种反演方法的两个主要特性是:将反射波至的振幅合并在一起进行分析,而且能得到没有人为干扰的(拾取走时曲线)精确结果。应用合成和野外资料的数值实例证明了我们反演程序的精确性。
We propose a two-step inversion procedure to choose subsurface velocity estimates from large offset seismic data. The first step is the automatic iteration of the large offset inversion method (ALLOIM), which we tested when there was a strong multiplicity of waves on the seafloor. Because the method relies on finding the location of the critical reflection point, the low-speed band can not be discerned, and reconstituting it is just an average estimate of speed. The second step in the inversion process is needed when there is a low velocity zone. This second step is a generalized least-squares method for subcritical reflections. We apply the perturbation technique to calculate the inversion solution and determine the parametric model from the average velocity estimate determined in the first step. The two main features of this inversion method are: combining the amplitudes of the reflected waves together for analysis and obtaining accurate results without human interference (picking up the time-of-flight curve). The numerical examples using synthetic and field data demonstrate the accuracy of our inversion procedure.