Abstract
The aim of this thesis is to investigate and gain better understanding of a new modeling technique called modeling by demigration. Demigration itself can be defined as the inverse of true amplitude migration. This involves nothing more than the formulation of a reflection imaging process by which one can return from a true amplitude depth migrated section to the original common offset section. Modeling by demigration represents a special implementation of the demigration concept, where the input is no longer standard migrated data but artificially migrated target geological structure(s) defined by the user. In this thesis the artificially migrated inputs are computed by employing both a standard/classical approach and the PSDM simulator approach of NORSAR (SimPLI technology).
A feasibility study has been carried out where the modeling by demigration concept has been compared with more standard modeling techniques based on dynamic ray tracing and Kirchhoff Helmholtz integral.
Synthetic data were generated for three different geological structures; a syncline, an anticline, and a fault. The output from the various modeling methods were compared based on both visual inspection as well as quantitative measures of relative amplitude ratios. In addition, the synthetic datasets were migrated to see how well the original geological structures could be mapped back.
The conclusions from this study were as follows;
The dynamic ray tracing method performed poor in the case of complex geology, as expected,
Modeling by demigration proved to be a feasible concept when benchmarked with the standard Kirchhoff Helmholtz modeling technique,
However, work needs to be carried out with respect to calibrations before direct (absolute) amplitude comparisons can be made, and
The SimPLI approach can represent an alternative to the standard artificial migration proposed in the original version of modeling by demigration.