Conclusions and future directions

The proposed prestack exploding-reflectors modeling preserves the velocity information in the starting prestack image. Taking advantage of the limited subsurface-offset range of partially focused images, several SODCIGs can be combined without degrading the velocity information necessary for MVA. My first tests of the method demonstrate that the number of independent experiments required for MVA can be substantially lower than the number of shot profiles in the original data set. They indicate that minimum number of independent experiments depends on the degree of focusing of the migrated image. The proposed method has the potential of significantly reducing the computational cost of MVA based on wavefield continuation migration and modeling, and to enable the use of expensive wavefields methods when estimating migration velocity.

The number of SODCIG that can be combined to form an independent experiment depends on the degree of correlation between the SODCIGs. In this paper I presented a simple schemes for combining SODCIGs that takes advantage of their limited offset range to achieve spatial decorrelation. However, temporal decorrelation can be also induced by modeling using spatially varying, and temporally decorrelated, source functions. This approach would follow the ideas used for phase encoding and Montecarlo migration. The additional degrees of freedom introduced by the variations in source functions are likely to enable further reduction of the number of independent experiments needed to preserve the velocity information necessary to improve velocity by MVA.