A new slant on seismic imaging: Migration and integral geometry
- 1 July 1987
- journal article
- Published by Society of Exploration Geophysicists in Geophysics
- Vol. 52 (7) , 943-964
- https://doi.org/10.1190/1.1442364
Abstract
A new approach to seismic migration formalizes the classical diffraction (or common‐tangent) stack by relating it to linearized seismic inversion and the generalized Radon transform. This approach recasts migration as the problem of reconstructing the earth’s acoustic scattering potential from its integrals over isochron surfaces. The theory rests on a solution of the wave equation with the geometrical‐optics Green function and an approximate inversion formula for the generalized Radon transform. The method can handle both complex velocity models and (nearly) arbitrary configurations of sources and receivers. In this general case, the method can be implemented as a weighted diffraction stack, with the weights determined by tracing rays from image points to the experiment’s sources and receivers. When tested on a finite‐difference simulation of a deviated‐well vertical seismic profile (a hybrid experiment which is difficult to treat with conventional wave‐equation methods), the algorithm accurately reconstructed faulted‐earth models. Analytical reconstruction formulas are derived from the general formula for zero‐offset and fixed‐offset surface experiments in which the background velocity is constant. The zero‐offset inversion formula resembles standard Kirchhoff migration. Our analysis provides a direct connection between the experimental setup (source and receiver positions, source wavelet, background velocity) and the spatial resolution of the reconstruction. Synthetic examples illustrate that the lateral resolution in seismic images is described well by the theory and is improved greatly by combining surface data and borehole data. The best resolution is obtained from a zero‐offset experiment that surrounds the region to be imaged.Keywords
This publication has 0 references indexed in Scilit: