Description du poste

This doctoral project will explore an alternative paradigm for modelling 3D geological architectures of subsurface with an improved formalism of concepts and uncertainties . This project aims at (1) improving the numerical formalisation of knowledge and hypotheses embedded in the subsurface representation and (2) improving the characterization of structural uncertainties.

This project will mainly address epistemic uncertainties, which partly result from the ambiguity and sparsity of available data. Ambiguity is generally compensated by manually introducing expert interpretive information (e.g., as interpretive cross-sections).

The student in charge of this project will develop an innovative approach that implements the cognitive interpretation process applied by geologists. Its formalisation will rely on:
examples gathered in a corpus of natural and simulated objects,
a formal description of geological concepts gathered in an ontology of structures, and
on an automated interpretation method based on spatiotemporal descriptors.

This new paradigm approaches geomodelling as an automated interpretation process based on artificial intelligence instead of a direct data interpolation. The proposed method will improve the integration of heterogeneous data, scale management, and exploration of epistemic uncertainties, while clarifying the structural concepts embedded in explored architectures.

For more information on the thesis, you can contact the supervisors :
Dr. Gautier Laurent : Maître de Conférences - ISTO [gautier.laurent[a]univ-orleans.fr]
Dr. Christelle Loiselet : Agent BRGM [c.loiselet[a]brgm.fr]
Dr. Yannick Branquet : Maître de Conférences - ISTO [yannick.branquet[a]univ-orleans.fr]