Global optimization for AVO inversion: A genetic algorithm using a table-based ray-theory algorithm

Amplitude Variation with Offset (AVO) inversion provides estimates of the P-wave velocity, S-wave velocity and density of a stratified medium. Global optimization is desirable for the inversion to account for the multi-parametric behavior of the AVO inversion which is strongly affected by the initial estimates of the model rock properties. We carried out an analysis to verify the dependency between P-wave, S-wave velocity and density in the recovered parameters using empirical relations as constraints. In inversion schemes, the forward modeling is often the most time-consuming process. To reduce computation time, we have implemented a genetic algorithm using a table-based ray-theory algorithm to allow for a large amount of vertically inhomogeneous models in the global search. Our results show that the genetic algorithm was capable of recovering the physical parameters with good agreement for examples using the empirical constraints. However, it sometimes converged to solutions which were far from the correct answer, but were good models to explain the observed dataset. The forward modeling algorithm has shown excellent performance to be used in global optimization schemes, because it allows the use of a large number of members in the population of the genetic algorithm.