Page 453 - Acquisition and Processing of Marine Seismic Data
P. 453

444                                 9. VELOCITY ANALYSIS




























           FIG. 9.17  Two semblance plots calculated from two synthetic CDP gathers with different folds including a 20% random
           noise. Semblance plots from (A) 120-fold CDP with a 12.5-m trace interval, and (B) 30-fold CDP with a 50-m trace interval.



              The effect of multiples on semblance plots is  rest of the enclosures in the semblance plot
           quite dramatic: the enclosures of primary reflec-  completely represent the amplitudes of the mul-
           tions on the semblance plots generally indicate a  tiples. Applying a suitable demultiple method
           velocity trend increasing with depth until the  may significantly improve the quality of the
           multiple amplitudes arrive. After that, the enclo-  semblance plots (Fig. 7.25).
           sures representing the multiple amplitudes
           become rather distinct on the semblance plots,  9.3.6 Effect of Time Gate
           producing concordant velocity trends almost
           parallel to each other for every multiple bounce  Velocity spectra are calculated over theoreti-
           (Fig. 9.19). In such cases, the processors must be  cal trial reflection hyperbolas computed using
           careful to pick velocities of only primary   a series of constant velocities. The semblance
           reflections.                                 values are obtained from the amplitudes on
              Particularly in shallow continental shelves,  the recorded CDP within a time gate along this
           seismic data may contain several orders of mul-  theoretical curve (Fig. 9.9). The time length of the
           tiple reflections, which makes the semblance  gate influences the quality of the resultant sem-
           plots rather complex. For instance, there are five  blance plot. If it is selected too short, it increases
           bounces of multiples visible on the CDP gather  the computational time of the semblance plot,
           in Fig. 9.20. The enclosures from the amplitudes  whereas the resolution along the time axis
           of these multiples are so dominant on the sem-  decreases when it is too large.
           blance plot that the enclosures corresponding   Fig. 9.21 shows semblance plots calculated
           to the primary reflections are restricted only  using different time gate lengths. As the gate
           between 300 and 600 ms zero-offset time. The  length increases, semblance enclosures become
   448   449   450   451   452   453   454   455   456   457   458