Page 537 - Acquisition and Processing of Marine Seismic Data
P. 537

528                                 11. SEISMIC MIGRATION

           obtain a distorted image of the subsurface using  plane, which moves the event along the orange
           our 2D seismic data. Even though we cannot   semicircular curve in Fig. 11.37C, where the
           solve any structures at some crossline offset to  amount of displacement (x) is indicated by
           the survey line with 2D datasets (e.g., the one  crossline displacement. This ambiguity can only
           at x distance away from the sail line in     be solved by recording 3D data and applying 3D
           Fig. 11.37B), they sometimes have strong influ-  imaging.
           ences on the data. Out-of-plane reflections can  As a result, 2D lines frequently involve side-
           only be properly focused after 3D migration.  sweeps as steep linear or curved reflections as
              The out-of-plane reflections are recorded as if  well as noncollapsed diffraction hyperbolas, even
           they are reflected from the 2D sail line plane at  after 2D migration (Fig. 11.37D). 3D migration
           the time they arrive at the receivers (green dot  revokes the assumption of 2D migration that
           in Fig. 11.37C), and during 2D migration, these  "all recorded data is reflected from the survey line
           amplitudes are treated as ordinary 2D reflec-  plane," and it displaces the events to their true
           tions recorded from directly beneath the seismic  subsurface locations in both inline and crossline
           line. Since the process is 2D, it is not possible to  directions. In Fig. 11.38, a 2D line (inline #150)
           move these reflections to their true reflected  from a 3D marine survey is processed by 2D post-
           positions, which are actually not located in the  stack time migration (Fig. 11.38A) to compare the
           survey line plane. In order to image the event  image quality with a 3D migration output
           at its true subsurface location (blue dot in  (Fig. 11.38B) of the same dataset. In the 3D
           Fig. 11.37C), it is necessary to back-propagate  migrated section, diffraction hyperbolas are bet-
           the event using 3D imaging along an oblique  ter collapsed, the events from the crossline































           FIG. 11.38  (A) A 2D line (inline #150) from a 3D survey processed by 2D poststack time migration. (B) The same line after
           3D migration. In the 3D migrated image, out-of-plane reflections are removed, fault planes are much clearer, and trace-by-
           trace consistency of the reflections is much better.
   532   533   534   535   536   537   538   539   540   541   542