Nowadays, methods of three-dimensional seismic analysis are part of every oil company's activity. The depth migration of 3-D seismic data has become mandatory, and interpreters feel more comfortable working on depth-converted images because of the complex geological structure of the subsurface of the oil fields under investigation. Any 3-D depth migration algorithm requires a velocity model in order to yield a depth image of the geology. The risk of drilling dry holes increases and the decision to implant production wells becomes difficult when that decision is based on the interpretation of images that have been depth-converted with an inaccurate, or imprecise, velocity model. An exact knowledge of the velocity is therefore required prior to any attempt to convert seismic data in depth.
Because the geologic structures of oil reservoirs are usually relatively local features, their mapping by seismic methods may become less reliable when rapid lateral velocity variations occur in the area. For this study, I make the hypothesis that the Dix assumption 1955 applies, but I know that this hypothesis is violated when I take into consideration the lateral variability on velocity.