Interferometric interpolation of missing seismic data
Interferometric interpolation of missing seismic data
复制标题
DOI：
10.1190/1.3110072
复制
发表时间：
2009-04
期刊：
影响因子：
3.3
通讯作者：
Yibo Wang;Yi Luo;G. Schuster
中科院分区：
文献类型：
--
作者：
Yibo Wang;Yi Luo;G. Schuster
我们开发了一种干涉插值方法来填充共射孔道的道隙。执行有两个步骤。首先，通过在镜头集合中交叉关联记录的轨迹，然后将所有镜头位置上的相关轨迹对相加，在轨迹间隙中创建虚拟轨迹。其次，将一维匹配滤波器应用于预测事件以校正小波和幅度畸变。通过一维假设（假设在共中点（CMP）集束中点周围有层状介质），干涉插值方法可以应用于CMP集束。它是一种完全数据驱动的方法，只需要原始数据。综合数据和现场数据的经验结果表明，近偏移迹线可以插值；但是由于有限的记录光圈所引入的近似，仍然存在一些伪影。为了比较，我们还使用了二维非平稳预测误差滤波器（PEF）来代替一维匹配滤波器。
We have developed an interferometric interpolation method for filling in trace gaps in common-shot gathers. There are two steps for implementation. First, virtual traces are created in the trace gaps by crosscorrelating recorded traces in a shot gather followed by summing correlated trace pairs over all shot positions. Second, a 1D matching filter is applied to the predicted events to correct for wavelet and amplitude distortions. By using a 1D assumption (assuming a layered medium around the midpoint in a common-midpoint (CMP) gather), the interferometric interpolation method can be applied to CMP gathers. It is a fully data-driven method that requires only the original data. Results with synthetic and field data empirically show that the near-offset traces can be interpolated; but because of the approximations introduced by a limited recording aperture, there are still some artifacts. For comparison, a 2D nonstationary prediction-error filter (PEF) was also used instead of a 1D matching filter in the se...