SKELETONS AND FRACTALS - A STATISTICAL APPROACH TO DEEP-CRUSTAL SEISMIC DATA-PROCESSING AND INTERPRETATION

Coherent seismic reflection events are identified using seismic skelet onization, an automatic pattern recognition technique in which wavefor ms are parameterized such that the attributes of each event are identi fied and stored in a relational database, or 'event file'. This new ap proach to seismic interpretation allows statistical analyses of the at tributes including estimation of scaling laws for reflectivity observe d on deep seismic data, and development of new filtering techniques ba sed on characteristics such as dip and reflection event length. Applic ations of the technique to deep crustal reflection data sets from the Alberta Basement Transect of the LITHOPROBE project assist interpretat ion in areas that lack surface outcrop or other means of relating refl ections from the deep crust to the surface. (C) 1998 Elsevier Science B.V. All rights reserved.