Processing the Teapot Dome Land 3D Survey with Madagascar

Surface Consistent Decon

The previous section, first look, created a few displays for the trace data, picked a predecon mute, a stack mute, and showed improvement from some basic processing. The 'first look' section used single trace deconvolution. Surface consistent decon (Hutchinson and Link, 1984; Cary and Lorentz, 1993) is more stable on data with poor signal and is widely used for land processing. Surface consistent decon computes a constant operator for each surface location.

Starting from the $RSFSRC/book/data/teapotdome/firstlook directory, you can look at the SConstruct file for the surface consistent decon with the commands:

cd ../scdeconpaper
gedit SConstruct

The key program for this processing stage is sftahscdecon. It was derived from sftahpef and is very similar. The Madagascar program allows a shot decon or a receiver decon to be applied. This is simpler than the multi component decomposition described in the references (Hutchinson and Link, 1984; Cary and Lorentz, 1993) and is an algorithm widely used in industry. The program computes a single operator for each ensemble of traces. To apply shot consistent decon, to data must be sorted to shot order and key='sx,sy' is input to sftahscdecon. The program averages trace autocorrelations and uses the header keys sx and sy to identify the end of an ensemble. At the end of the ensemble the averaged autocorrelation is used to design a prediction error filter (decon filter) and applies the filter to all traces in the gather. After shot decon the data is written to a file so the data can be sorted into receiver gathers and sftahscdecon used to apply receiver consistent decon. Other than the key parameter, the parameters in sftahscdecon are the same as the parameters in sftahpef. This SConstruct file used the same parameter values for the length of the decon operator and the decon design gate (minlag, maxlag, minfor, and maxcorr.) as used in sftahpef in the firstlook directory. Since decon will be applied twice in the surface consistent processing, pnoise was increased from .01 to .1 to prevent “over whitening” the data.

Figure 11 shows a few CDP gathers with shot and receiver consistent decon applied. Figure 12 shows the stack for line 141 with surface consistent decon.

scdeconcdps
Figure 11. A few CDP gathers with shot and receiver consistent decon applied.

scdeconstack141
Figure 12. The stack for line 141 with shot and receiver consistent decon applied.

Processing the Teapot Dome Land 3D Survey with Madagascar