Non-linear Deconvolution Based On Blocky Geological Structure Constraint

Seismic data resolution and SNR(signal to noise ratio)are of vital importance to investigate geologic structure and detect subsurface mineral resources.Deconvolution plays an important role for enhancing the seismic data resolution,but traditional deconvolution methods,based on wiener filtering,cannot break through the limit of the effective frequency band for the seismic signal,which can't meet the geological requirements interpreting the high resolution seismic data.It has been a new research direction for deconvolution to enhance the seismic data resolution significantly by introducing novel regularization constraint in the framework of nonlinear theory.Attenuation caused by viscoelastic medium not only reduces the seismic data resolution,but leads to non-stationary wavelets,restricting application effect of the traditional deconvolution methods based on the assumption of stationary wavelet.Introducing the time-varying wavelets into deconvolution,non-stationary sparse deconvolution could recover the strata reflection coefficient better.However,although improving the vertical resolution of seismic data,the single-channel sparse deconvolution reduces the lateral continuity of seismic profiles owing to lacking lateral constraint of geologic structure.Based on the previous research,for the problem existing in the single-channel sparse deconvolution,we proposed a novel deconvolution introducing blocky geological structure constraint.From the experiment,it indicates that our method could suppress the noise from the seismic data better,and recover the lateral variation characteristics of the reflection coefficient in the meanwhile enhance the seismic data resolution extremely.