Increasing the Velocity Spectra Resolution of Seismic Data Using Boot Strapped Method

Semblance analysis is one of the common methods in the seismic velocity estimation, which is based on the summation of samples in the velocity analysis window. In the case of deep reflection events or low offset data, the velocity spectrum of this method has a low resolution. An increase in the resolution of the semblance results in an improvement in the accuracy of the estimated velocity for NMO correction and stacking. In this paper, the bootstrapped statistical method was introduced to improve the resolution of velocity spectra. In semblance method, if the selected velocity for NMO correction is slightly different than the actual velocity, the traces inside the analysis window are not fully horizontal. Nevertheless, due to the small time difference between the adjacent traces, the semblance value calculated for this case is nearly identical to the value of the actual velocity and causes the extension of the peak of the velocity spectrum. If the velocity of the analysis window is different from the correct value, the bootstrapped method increases the time difference of two adjacent seismic traces by displacing them. By defining a new measure to compute semblance, this procedure can be useful to estimate the correct velocity and increase the resolution of seismic velocity spectra. The efficiency of this method is evaluated by applying it to both synthetic and real seismic data and it is compared with the result of semblance method. The obtained results show that the velocity spectrum of the bootstrapped method has a better resolution than that of the conventional semblance method.