利用多元地震属性预测储层信息

地震属性分析技术在储层横向预测中发挥着越来越重要的作用。利用多元地震属性分析技术进行储层信息预测的理论基础是，储层中岩石性质、流体性质的空间变化可引起地震反射波形、振幅、频率等地震属性的变化。本文主要阐述了利用地震属性的分类(聚类)分析和多元地震属性回归技术，来分析特定区域储层纵、横向变化特征；强调在区域地质概念模型的前提下，预测储层的分布范围和储层物性。应用实例表明，本文方法的预测结果与钻井资料吻合较好。

Using multivariate seismic attributions to predict reservoir information

Seismic attribution analysis technique plays a more and more important role in lateral reservoir prediction.Using multivariate seismic attribution analys technique for prediction of reservoir information is mainly based on spatial variation of rock characters and fluid properties in reservoir,resulted in variation of seismic attributions such as seismic reflected waveform,amplitude and frequency etc..The paper mainly described that using classification (cluster) analysis and multivariate seismic attribution regression technique to analyze the vertical and lateral reservoir variation characters in special areas,and puts stress on the predicted reservoir distribution scope and physical properties of reservoir in a model of regional geologic concept.The real cases showed that the predicted results by the method well coincides with drilling data.