地震信号内插与噪音剔除（一）

信号内插与噪音剔除是相互有着有机联系的两个方面。首先讨论地震信号的内插方法,提出了一种“最佳内插算子”,从而导出了另一种“检噪算子”,它可以有效地把干扰波从记录中识别出来,并将其“剔除”出去。进而讨论空间域数据的内插问题--道内插问题。在一定的条件下,空间域数据内插与时间域内插是完全等效的,因此道内插也可以采用与时间域相同的“最佳内插算子”。地震记录上的干扰波其绝大部分表现为空间域的脉冲式干扰,所以用“检噪算子”在空间域作褶积,便能识别干扰,从而对干扰加以剔除。本文解决了一种迭代的剔除方法,取得了好的效果。

SEISMIC SIGNAL INTERPOLATION AND NOISE DELETION

Seismic signal interpolation and noise deletion are two problems closely related each other. In this paper the author found some Optimum Interpolating operators at first, and from them, several Noise Detecting operators were derived. We can separate the noises from a seismic record and delete them.Author points out that there exists an identity between time domain interpolation and space domain interpolation. So we can use those time domain operators to perform Trace Interpolation effectively.On the other hand, we can use those time domain noise detecting operators delete the noises on a seismic record in the space domain.Note that most seismic noises appear as spike-like bursts while we rearrange the data samples in trace order. So noise detecting operator car recognize the noise spikes and delete them. The deletion process is an iteated method using subtraction after convolution (noise detection).This method may solve the problem of those very low S/N field records in data processing Also, it can help to avoid the 'SMILING' effect in migration process.This paper has two parts. In part II, author will extend this method to two dimensional interpolation and 2-D noise deletion.