最小展布子波反褶积

MINIMUM SPREAD WAVELET DECONVOLUTION

地质矿产部第一海洋地质调查大队

The 1st Party of Party of Marine Geological Surtvey, Ministry of Geology and Mineral Resources )

本文应用Backus-Gilbert线性反演理论的最小展布原则,对褶积地震模型在保持源子波振幅谱面积不变的约束条件下,推导了一种最小展布子波反褶积方法。由约束条件给出最小展布意义下的整形算子,克服了最小平方反褶积中期望子波人为选择的随意性。理论模型研究结果表明,该方法能较有效地压缩子波旁瓣和持续时间,提高子波的分辨力。进一步将该方法应用于实际资料处理,也取得了明显的效果。处理质量比常规脉冲反褶有较大的改善。

According to the Minimum Spread Principle of the Backus- Gilbert linear inversion, a new technique ofdeconvolution——the minimum spread wavelet deconvolution, was developed. Its application should be under thecondition that the area of the source wavelet amplitude spectrum must be invariable for the model of convolution. The snapping factor was given under the constraint of minimum spread,hence, the determination of desired wavelet from the least square deconvolulion was less dependent on the artifical factors. Results from theoretical model show that the sidelobe of the resolving’kernal could be surpressed and the duration of the Vavelet could also be shortened. As a result, the resolution of the wavelet was obviously improved. Its application to the actual data shows that it works better than the least square deconvolulion in improving the resolution and the continuity of the reflection events.