Eigenimage analysis of shallow seismic sections

The appearance of seismic events related to various subsurface objects is to a large extent defined by their coherency on seismic sections. In particular, shallow subsurface inhomogeneities can be expected to appear as the areas of poor coherency. Identifying and separating coherent and non-coherent events on the sections can be performed using the technique of eigenimage analysis. This technique is based on the singular value decomposition (SVD) of the sections represented as data matrices. A similar approach is implemented in the Karhunen-Loeve (or principal component) reconstruction method (Hemon and Mace, 1978). The technique has been successfully applied to various geophysical problems (Jones and Levy, 1987; Freire and Ulrych, 1988; Jackson et al., 1991; de Franco and Musacchio, 2001).