The Method Of Q-estimation Based On Pre-stack CMP Gathers

In recent years,the seismic shift from traditional structural exploration to lithological exploration.In the stage of lithological exploration,the urgent problems include reservoir prediction,crack detection,analysis of lithology and fluid properties,and so on.Quality factor value is closely related to subsurface information.It can be used for identification and analysis of seismic attenuation and frequency dispersion,improving reservoir characterization and the accuracy of reservoir prediction.To estimate seismic wave attenuation accurately can enhance the resolution of seismic imaging and the reliability of crack detection.Therefore,Q value is always one of the key topics in the field of non-elastic medium research.At the beginning of this article,we discuss the theory of viscoelastic medium and seismic wave attenuation mechanism and its influencing factors,and introduce the application of time-frequency analysis methods and their application conditions.Then,we discuss several conventional methods of Q estimation,compare the advantages and disadvantages of peak frequency method,frequency shift method and spectral ratio method,study the Q simulation of seismic wave attenuation based on time-varying convolution model.Q is a useful physical parameter and rarely been accurately measured.The Q estimation based on traditional post-stack seismic data wasn't accurate and its application wasn't very well.Besides,the Q estimation based on VSP data is accurate,but their data coverage is too small to be used extensively.In order to overcome these difficult,we present a method of Q estimation based on pre-stack CMP gathers in this paper.The time-frequency resolution of generalized s-transform can be adjusted in this method.We take this generalized s-transform to do the time-frequency analysis of pre-stack CMP gathers.The time-window of generalized s-transform can vary with frequency regulation,which reduces the spatial spectrum caused by the thin layer interference.And the amplitude distortion correction eliminates the non-hyperbolic NMO and NMO stretch effects.For practical data,to improve the SIN ratio,corrected complex spectra are stacked over an offset-range window.Select a suitable reference axis,for spectral ratio logarithmic varying linearly with frequency,we can obtain the slope of spectral ratio.The variation of apparent Q versus the square of offset(QVO)is extrapolated to give a zero-offset Q estimation.The interval Q-factor is estimated from CMP gathers using a layer-stripping approach.A byproduct of this inversion is AVO information without attenuation.Application of model data and actual data have achieved good results,proving that Q estimation method based on pre-stack CMP gathers is simple,effective and accurate,and can be used for hydrocarbon detection.