next up previous print clean
Next: Stacked sections Up: Multiple attenuation results Previous: Outside the salt boundaries

Inside the salt boundaries

Figure [*] displays the multiple attenuation results for the second CMP gather in Figure [*]a. The multiple attenuation becomes much more challenging because of various 3-D effects not taken into account in the multiple modeling step. The HRT result in Figure [*]b shows a lot of multiples for which the moveouts are not hyperbolic, thus not well attenuated in the Radon domain. The adaptive subtraction result in Figure [*]a indicates that the modeling inaccuracies at short offsets tend to degrade the estimated signal. The pattern-based method with 2-D PEFs in Figure [*]c does better than the adaptive subtraction technique at both short offsets and late times. The adaptive subtraction technique does better than the 2-D PEFs in the middle of the CMP gather. Again, the best multiple attenuation results are obtained with the pattern-based method with 3-D PEFs (Figure [*]d). Note that a strong primary at 3.7 s. is slightly attenuated with the pattern-based approach. For this event, primaries and multiples might be correlated across offset and shot position. A bigger ${\epsilon}$ in equation (7) would solve the problem by allowing more energy from the signal to be preserved.

 
comp-multatt-cmp.15500
comp-multatt-cmp.15500
Figure 5
Multiple attenuation results for the CMP gather in Figure [*]. (a) Signal estimated after adaptive subtraction. (b) Signal estimated after HRT and mute. (c) Signal estimated with a pattern-based method and 2-D PEFs. (d) Signal estimated with a pattern-based method and 3-D PEFs.
[*] view burn build edit restore


next up previous print clean
Next: Stacked sections Up: Multiple attenuation results Previous: Outside the salt boundaries
Stanford Exploration Project
7/8/2003