next up previous print clean
Next: Outside the salt boundaries Up: Guitton: Multiple attenuation Previous: Multiple attenuation with a

Multiple attenuation results

Now, I show the multiple attenuation results obtained with field data from the Gulf of Mexico (Figure [*]). This dataset corresponds to a deep water survey (water bottom below one second) with a shallow salt body, thus creating strong multiples. We have only one streamer for this dataset making the modeling of 3-D multiples impossible. The three methods detailed in the preceding sections are used to attenuate the multiples.

Note that the Delft and pattern-based approaches are carried out in the shot domain, as opposed to the CMP domain for the HRT. Thus I decided to make the comparison of the three techniques in the CMP domain. Figures [*] and [*] show two CMP gathers along with the multiple model computed with the autoconvolution of the shot gathers. One CMP gather is located outside the salt boundaries, i.e., in the sediments (Figure [*]). The other gather is located inside the salt boundaries (Figure [*]). These two areas allow us to see the strengths and weaknesses of each method for different environment.

 
compstack-datmult
compstack-datmult
Figure 1
(a) Stacked section of the input data with multiples. (b) Stacked section of the multiple model obtained with the Delft approach. Note that the kinematics of the multiples is very accurate.
[*] view burn build edit restore

 
comp-datmult-cmp.3200
comp-datmult-cmp.3200
Figure 2
CMP gathers extracted at midpoint location 3200 m. (a) CMP gather of the data with multiples. (b) Multiple model. Note that the near offset traces have the wrong amplitudes.
[*] view burn build edit restore

 
comp-datmult-cmp.15500
comp-datmult-cmp.15500
Figure 3
CMP gathers extracted at midpoint location 15500 m. (a) CMP gather of the data with multiples. (b) Multiple model. Note that the near offset traces have the wrong amplitudes.
[*] view burn build edit restore



 
next up previous print clean
Next: Outside the salt boundaries Up: Guitton: Multiple attenuation Previous: Multiple attenuation with a
Stanford Exploration Project
7/8/2003