Page 288 - Acquisition and Processing of Marine Seismic Data
P. 288

5.6 GAIN RECOVERY                              279

































                     P
           FIG. 5.39  t corrections for different P values on a marine shot gather from relatively deeper waters. (A) Filtered shot with
                       P
           no gain, and its t corrected versions for (B) P ¼ 1.0, (C) P ¼ 2.0, (D) P ¼ 4.0, (E) P ¼ 6.0, and (F) P ¼ 10.0.
           gathers from a shallow continental shelf and a rel-  window. Fig. 5.40 shows the principle of AGC
           atively deep basin area, respectively. Selection of  application. A window with a length of Δt is
           the proper P value depends on the recording  selected (Fig. 5.40A), and this window is pro-
           length and seabed depth. In general, smaller P  gressively moved down along the time axis
           values are appropriate for the seismic data from  sample-by-sample (e.g., Fig. 5.40B and C). Here,
           shallow shelf areas while higher values are  Δt is also known as AGC operator length. Each
           required for seismic data from deeper waters, to  time, a scale factor is calculated using the ampli-
           properly recover the amplitudes of late arrivals.  tudes within the specified time window. This
           For instance, P ¼ 2.0 is suitable for the shot in  scalar may equal to the inverse of root mean
           Fig. 5.38,but P ¼ 4oreven P ¼ 6can be selected  square (RMS), mean or median amplitude value
           fortheshotinFig.5.39.Ingeneral,ifalowerPvalue  of the time window. Although the scalar can be
           is selected, then the amplitudes of deep reflections  multiplied by the sample value located at the
           arenotsufficientlyrecovered.Onthecontrary,ifan  center, beginning or end of the AGC window,
           unnecessarily high P value is used, then the ampli-  generally the amplitude value at the center of
           tudes of early arrivals almost die (Fig. 5.38F).  the window is preferred, as is the case in
                                                        Fig. 5.40. Then the window is moved one sample
                                                        and a new scalar is calculated and applied to the
           5.6.3 Automatic Gain Control
                                                        central sample of the window. The AGC opera-
              Automatic gain control (AGC) is one of the  tion continues in this fashion along the time axis
           most common gain recovery methods in seismic  from the beginning to the end of the time sam-
           processing. AGC is applied to the seismic data  ples within the trace. At the beginning and
           on a trace-by-trace basis using a sliding time  end of the trace, however, there will not be
   283   284   285   286   287   288   289   290   291   292   293