Time-frequency analysis of deep crustal reflection seismic data using Wigner-Ville distributions

One important component of deep crustal reflection seismic data in the abse nce of drill-hole data and surface-outcrop constraints is classifying and q uantifying reflectivity patterns. One approach to this component uses a rec ently developed data-decomposition technique, seismic skeletonization. Skel etonized coherent events and their attributes are identified and stored in a relational database, allowing easy visualization and parameterization of the reflected wavefield. Because one useful attribute, the instantaneous fr equency, is difficult to derive within the current framework of skeletoniza tion, time-frequency analysis and a new method, empirical mode skeletonizat ion, are used to derive it. Other attributes related to time-frequency anal ysis that can be derived from the methods can be used for shallow and deep reflection seismic interpretation and can supplement the seismic attributes accrued from seismic skeletonization. Bright reflections observed from bel ow the sedimentary basin in the Southern Alberta Lithosphere Transect have recently been interpreted to be caused by highly reflective sills. Time-fre quency analysis of one of these reflections shows the lateral variation of energy with instantaneous frequency for any given time and the lateral vari ation of energy with time for any instantaneous frequency. Results from emp irical mode skeletonization for the same segment of data illustrate the dif ferences in the instantaneous frequencies among the intrinsic modes of the data. Thus, time-frequency distribution of amplitude or energy for any sign al may be a good indicator of compositional differences that can vary from one location to another.