Optimizing full stacked PP-PS seismic data interpretation

This paper optimizes the interpretation of a full stacked PP and PS data. The optimization is done in two important stages that are PP time PS registration and the data inversion. An enhanced major event analysis is developed to optimize the PS to PP time registration. It contains an automatic event picking and objective function minimization. That is a simple technique, however fast and provides good PS to PP data alignment. An innovative stochastic inversion is then used to estimate the compressional (P), share (S) impedance and Vp/Vs ratio from the PP-PS data set. The inversion uses seismic attributes, both in time and frequency domain, and locally varying anisotropy (LVA) to build prior model and to enhance the selection criteria. Iterative low frequency model inclusion is implemented during inversion process. Using more constraint from seismic attributes and iterative low frequency model inclusion, the inversion yields high detail properties in all over the data volume with highly consistence to well log and seismic data. The availability of PS data enhances the full stacked seismic data inversion to provide share impedance as well as Vp/Vs ratio properties which are very useful for not only reconstructing the reservoir structure but also predicting the lithology and pore fluid. © 2016 SEG.