Prestack Seismic Inversion via Nonconvex L1-2 Regularization

Using seismic data, logging information, geological interpretation and petrophysical it is possible to estimate the stratigraphic texture elastic parameters of a study area via inversion. As such, inversion an indispensable tool in field oil gas exploration development. However, due unknown natural factors, inversions are often ill-conditioned problems. One way work around this unknowable information determine solution using regularization methods after adding further priori constraints. In study, nonconvex L1−2 method innovatively applied three-parameter prestack amplitude variation angle (AVA) A forward model first derived based on Fatti approximate formula then low-frequency models for P impedance, S density established horizon data. Bayesian framework, we derive objective function AVA To improve accuracy stability results, remove correlations between that act as initial constraints Then, solved by method. Finally, validate our applying synthetic observational data sets. The results show yields highly accurate, laterally continuous, can be used identify locate reservoir formation boundaries. Overall, will useful future focused predicting location reservoirs.