Full text loading...
-
Elastic Anisotropic Full Waveform Inversion Using Probabilistic Petrophysical Constraints
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, 82nd EAGE Annual Conference & Exhibition, Oct 2021, Volume 2021, p.1 - 5
Abstract
Full waveform inversion (FWI) works by iteratively minimizing an objective function that measures the misfit between observed and predicted data in the least-squares sense. However, FWI suffers from significant problems. First, the inversion solved by gradient techniques may not lead to the globally optimal solution. Second, all wave propagation mechanisms are not adequately considered if one does not assume a stiffness tensor structure that truly represents the subsurface. Third, depending on the parameterization used for inversion, elastic properties may be coupled and updates of one parameter may impact others, an effect known as interparameter crosstalk. Additionally, some combinations of model parameters can be lithologically implausible, and not represent feasible lithological units. We derive anisotropic subsurface models using elastic FWI and explicitly impose petrophysical penalties to recover models consistent with the seismic data as well as with the petrophysical context in the area. This methodology reduces the potential negative impact of local minima, mitigates interparameter crosstalk artifacts, and avoids geologically implausible models. We define this penalty using probability density functions derived from petrophysical information. The proposed FWI objective function leads to robust anisotropic models that represent plausible lithologies, while at the same time leading to data predictions consistent with the observations.