The Research Of Velocity Analysis Methods In Prestack Time Domain

In this paper, we mainly study the ellipse evolving stacking velocity analysis method andprestack time migration common scattering point gathers velocity analysis method in prestacktime domain.Firstly, we study the defect in conventional velocity analysis method. And then theprinciple of ellipse evolving stacking velocity analysis method is introduced by a singlecommon reflection point model. The reason that the ellipse evolving stacking velocityanalysis method can get more accurate velocity is introduced too. The velocity analysis flowof ellipse evolving stacking velocity analysis method is given as a researching result that isdifferent from the previous researcher’s method which is intervening stacking. It contains fourparts in the velocity analysis flow.(1)how to select the gathers in ellipse evolving, and theexpression to select gathers is given in this paper;(2)the flow of generating stacking velocityspectra using ellipse evolving method is also given;(3)studying the convergence methods ingenerating velocity spectra. Comparison with four convergence methods, one of them,normalization cross-correlation sum algorithm, is good for ellipse evolving velocity spectraformation; another method that is statistical phase correlation algorithm can also get a goodspectra; in this paper, we put forward a best method that is statistical phase correlationalgorithm after filtering combining with normalization cross-correlation sumalgorithm;(4)interactive picking-up based on GUI platform in MATLAB is achieved. Ellipseevolving stacking velocity analysis method can get more accurate velocity than conventionalCMP method by testing on typical geological model and2D real data. When signal-to-noiseequals one, ellipse evolving stacking velocity analysis method is also good.By introducing equivalent offset, we can obtain the single square root travel-timeequation from the double square root travel-time equation. Common-scatter-point(CSP)gathers extracted and built from the single square root travel-time equation have the advantages of gaining the number of coverage and spatial anti-aliasing. We can do velocityanalysis using CSP gathers and get prestack time migration section after stacking them. Thismethod is effective by testing on model and real data.But there are mapping noise in theprocess of generating CSP gathers. Mapping noise may reduce the accuracy of pickingvelocity and the quality of migration. On this basis, I study the producing mechanism of themapping noise in the CSP gathers and suppress them by introducing amplitude correctionfactor. With testing on model and real data, we can see that the spectrua and common scatterpoint gathers are improved greatly after noise suppression, and the quality of migration is alsoimproved obviously.In the end, the ellipse evolving stacking velocity analysis method and prestack timemigration common scattering point gathers velocity analysis method are applied in the3-dimension seismic data, and the two methods are achieved by parallel computing.