Spectral Re-composition And Its Application In Reservoir Prediction

The exploration of oil-gas field becomes more and more mature,and the field of such exploration becomes bigger,so the difficulty of the exploration improves.To provide abundant and efficient data for oil field exploration,we must extract much more hidden information which can't be caught out in seismic data,then,seismic attribute analysis satisfies with this target.Spectral re-composition based on least square inversion assumes that the spectrum of seismic data makes up with Ricker wavelets as its components.In this paper,using least square inversion iteration,with conjugate gradient method to solve the nonlinear equations,then,re-compositing the Ricker wavelet component's dominant frequency and its amplitude to receive the Ricker wavelet component.For different frequency has different sensibilities to different geology structures,we apply the re-composition to two aspects:One is to extract horizon slice from 3D seismic data,then re-compositing spectrum in every single trace,using the achieved dominant frequency and amplitude as two attributes to re-composite the dominant frequency attribute and amplitude attribute,applying them to seismic attribute analysis.Then using achieved two new attributes to make cluster analysis,comparing with other seismic attributes,this method has better result.Another is to calculate the mean spectrum of 2D section,simultaneouslycalculate the Ricker wavelet components,using such components with different frequency as filter to filter the seismic data,after filtering,structures in deep layer becomes much more obvious.For finding out the relationship between attribute extracted,we use the principal component analysis to optimize the seismic attribute.