Depth remigration by means of the image wave equation

Author(s)

Fernando Perin Munerato

• Document type: Master's Dissertation
• Press: Campinas, SP.
• Institution: Universidade Estadual de Campinas (UNICAMP). Instituto de Matemática, Estatística e Computação Científica
• Defense date: 2006-03-31

Examining board members

Joerg Dietrich Wilhelm Schleicher; Ricardo Caetano Azevedo Biloti; Mauricio Boaventura

• Advisor: Maria Amélia Novais Schleicher; Joerg Dietrich Wilhelm Schleicher

Abstract

This work approaches the question of how to solve the image-wave equation for depth remigration by numerical methods. The objective is the reconstruction of an image of the geologic layers of the subsoil from a previously migrated image with a different velocity model. Our main objective in this work is the investigation of possible methods that can solve the problems that appeared when using explicit _nite-difference schemes for the solution of the image-wave equation in previous works, particularly numerical dispersion. For this purpose, we study the method of _nite volumes, as well as implicit _nite-difference schemes. The main characteristic of the _nite-volume method is to simply propagate the averages in the cells of the mesh instead of the discretized data themselves as it is done in the _nitedifference method. As another attempt to solve the problem of the dispersion, we study two types of implementation of implicit _nite-difference schemes, that is, implicit implementations of conventional schemes evaluated out the edge of the cell and a scheme evaluated in the center of the cell. The quality of the studied algoritms has been tested numerically. These numerical tests show that the method of _nite volumes is not adequate to solve the problem of dispersion, for the average calculated in each step additionally increases the pulse stretch. Moreover, the implicit implementations of the conventional schemes show the same dispersion behavior as the explicit implementations. Solely the centered scheme was capable to improve the numerical dispersion in comparison with the previous implementations, however only for data containing (AU)