next up previous print clean
Next: FUTURE WORK Up: INTERPOLATING MISSING TRACES Previous: INTERPOLATING MISSING TRACES

Example

Figure interpIn shows a cube of input data. It is a window from several shot gathers, reversed and added to itself to produce data with both seismic wavefield character and a complex set of conflicting dips. The upper panel shows the original cube, the lower panel shows the same cube after zeroing every other trace.

Figure smoothInterp shows the same data after every other trace was zeroed, and then re-estimated with smoothly varying PEFs. The lower panel of the figure shows the residual. For comparison, Figure patchInterp shows the same data after interpolation with a patching algorithm, along with the residual in the lower panel. The patching result is not bad, but it does not have the accuracy of the smoothly varying result.

 
interpIn
interpIn
Figure 2
An interpolation test data cube. The top panel is the original data cube; the bottom panel is the same data after every other trace was zeroed. The bottom panel is the input to the interpolation.
view burn build edit restore

 
smoothInterp
smoothInterp
Figure 3
Output of smooth interpolation. Top panel shows interpolated data, bottom panel shows residual. Compared to Figure patchInterp, there is little energy in the residual and it is difficult to tell which traces are interpolated. Clips are set equal.
view burn build edit restore

 
patchInterp
patchInterp
Figure 4
For comparison, output of patched interpolation. Top panel shows interpolated data, bottom panel shows residual. The residual is large compared to Figure smoothInterp. Clips are set equal.
view burn build edit restore


next up previous print clean
Next: FUTURE WORK Up: INTERPOLATING MISSING TRACES Previous: INTERPOLATING MISSING TRACES
Stanford Exploration Project
7/5/1998