Strategies for the application of a conjugate gradient FWI algorithm of shallow environments

Abstract

In this work we evaluate different strategies for the model update inside a FWI algorithm. It is know that the S wave velocity parameter is the one with greater sensitivity in a data covered by surface waves. The studied strategies, however, attempt to recover the P wave velocity and the density models in different ways. We compared the pure optimization through a conjugate gradient algorithm with the direct calculation assuming a constant Poisson ratio and by Gardner equation. The results show good performance of the both methodologies in simple 1D models. More research is necessary in more complex situations.