Non-Linear Cross-Equalization of Time-Lapse Seismic Surveys Using Artificial Neural Networks

P035 NON-LINEAR CROSS-EQUALIZATION OF TIME- LAPSE SEISMIC SURVEYS USING ARTIFICIAL NEURAL NETWORKS Summary 4D time-lapse seismic interpretations are hindered by the lack of repeatability between the two different time-lapse surveys especially those including legacy 3D surveys or those acquired with different shooting geometry. Artificial neural networks can provide a means of cross-equalizing such surveys. Their ability to act as non-linear mapping functions can overcome the limitations of other linear algorithms and can operate prestack when acquisition geometries do not match. A test case is performed on two time-lapse surveys acquired over a North Sea oil field in 1989 and 1992.