FRACTALS IN DEEP-CRUSTAL SEISMIC-REFLECTION PROFILES

Application of a syntactic pattern recognition technique, seismic skel etonization, to deep crustal seismic reflection data allows attributes such as energies, lengths and dips to be associated with individual r eflection events. Some of these attributes exhibit fractal properties, e.g. the relationship between seismic event lengths and their spatial distribution throughout the crust. This approach provides a new techn ique to analyse complex geometry on seismic reflection data.