信噪比谱约束下的井控反褶积方法研究

Study on Well-controlling Deconvolution Constrained by S/N Spectrum

基于“反射系数白噪。子波最小相位”假设的常规叠后反褶积方法．在提高地震资料分辨率的同时往往会降低资料的信噪比，不利于薄储层预测。文中通过引入测井资料的高频信息，对常规反褶积方法进行改进，确保了高频信息的真实性，实现了井控反褶积方法。通过分析当前求取信噪比的几种方法．最终选择互相关法计算地震资料的信噪比，并选用三角滤波器分频求取地震资料的信噪比得到信噪比谱。对比反褶积前后地震资料的信噪比谱，认为约束井控反褶积参数选取，确保反褶积结果提高分辨率的同时保持高信噪比。将该方法应用于实际地震资料，结果表明井控反褶积后目的层段地震资料频谱得到一定的拓展，有效带宽内信噪比基本保持不变，并略有提高，为提高薄层的识别能力奠定了资料基础。

While the conventional deconvolution with post-stack based on the hypothesis of white random noises as reflection coefficient and wavelet with minimum phase can enhance the seismic data resolution, the signal-to-noise ratio is usually decreased, unfavorable for the prediction of thin reservoirs. With the high frequency information of well logging data introduced, the conventional deconvolution method is improved to ensure the true information at the high frequency end, realizing the implementation of the well controlling deconvolution. After analysis of several methods currently used for S/N ratio calculation, the correlation method is chosen finally to compute the signal-to-noise ratios of seismic data and a splitting frequency method with triangular filter is used to get the signal-to-noise spectrum. By means of comparison of the S/N ratio spectrums of previous seismic data and post deconvolution, the selection of well controlling deconvolution parameters is constrained, ensuring high resolution as well as keeping high S/N ratio. The application of the method in real seismic data has shown that the spectrum band of seismic data at the target formations after well controlling deconvolution has been extended somewhat and the signal-to-noise ratio in the effective spectrum band keeps basically constant, with slightly increase, providing information basis for the enhancement of thin stratum identification.