同态解卷在地震测深中的应用

Application of Homomorphic Deconvolution to Deep Seismic Sounding

在地震测深资料的处理分析中,震相的识别是至关重要的一步,只有震相识别正确,后续的一系列反演工作才能顺利进行。而目前震相识别在国内外地震测深中仍是一个令人困惑的问题。 本文将同态解卷技术应用到地震测深中,从一个新的角度开展了工作。首先用功率谱滤波法对记录进行预处理,降低噪音的影响。而后,根据同态解卷对震源子波估计的有效性,提出了多道时域迭加法,使子波估计更加真实可靠。在此基础上,利用震相间子波的相似性进行震相识别。理论模拟及实际资料处理的结果表明:该法是行之有效的,它为震相识别提供了一条新的途径。

In the analysis of deep seismic sounding how to reduce the noise effect and to recognize seismic phase has been a puzzled problem.This paper applies homomorphic deconvolution technique to the seismic sounding. Firstly, it pre-processes the recording data by using power spectrum filtering and then it presents a multichannel time -stacking method to make the wavelet estimation more reliable. Finally, according to the similarity of seismic phase wavelet, a new way is proposed to recognize seismic phase. The results of simulation and real data processing have shown that this new method is very useful.