FWI sensitivity analysis in the presence of free-surface multiples
V.V. Kazei
B.M. Kashtan
V.N. Troyan
W.A. Mulder
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
It is generally believed that full waveform inversion needs very low frequencies in the data to avoid convergence to a local minimum, which would lead to an incorrect velocity model. Often, low frequencies are not present in the data. Then, a kinematically accurate initial model is required that contains the long-wavelength structures that cannot be reconstructed from the data. Mora (1989), however, showed that reflector below the target area may help in recovering some of the longwavelength information. Here, we extend his analysis to the case with a free surface, generating multiples. At first sight, multiples should improve the sensitivity and resolution. Indeed, our study shows that sensitivity becomes higher in some parts of the model spectrum when multiples are included. If we consider thin layers and only first-order multiples, then the sensitivity is improved. However, when the layer is thick enough to allow for normal modes of higher orders, areas of high sensitivity appear and this leads to a poorer conditioning of the inverse problem in the wavenumber domain.