Fast geoelectrical modeling and imaging based on multigrid quasi-linear approximation and electromagnetic migration

In recent years, interest has been markedly increasing in the electromagnetic method for offshore petroleum exploration, especially for the method which uses a controlled electromagnetic source (the marine controlled source electromagnetic (MCSEM) method). The interpretation of MCSEM data requires numerous three-dimensional (3-D) modelings of the responses in the receivers for different geoelectrical models of complex sea-bottom geological formations. This task may be extremely expensive even on modern computers. In order to overcome this significant difficulty, a new approach to both forward modeling and inversion of MCSEM data is introduced in this dissertation.; For fast modeling, an application of the quasi-linear (QL) approximation for the integral equation (IE) method in the framework of the multigrid method is introduced, and a capability of fast modeling without losing accuracy especially for MCSEM data is demonstrated.; As a new effective inversion, novel ideas of electromagnetic (EM) migration are applied to the interpretation of a typical MCSEM survey and are successfully developed. The principles of the migration imaging and iterative migration are formulated and implemented in the corresponding computer code.; The developed modeling and inversion schemes have been tested using synthetic problems and practical data collected at the Gemini prospect, the Gulf of Mexico, and in the Troll West Gas Province of the North Sea. The results demonstrate that the MGQL approximation provides a fast and accurate solution for numerical modeling of the multitransmitter EM data in complex 3-D geoelectrical structures with lesser cost than a standard IE solution. The results of EM migration inversion show that migration can be treated as an efficient prospective method of MCSEM data interpretation.