Modeling of anisotropic ice flow in Elmer/Ice using machine learning (M/F)

Cnrs Inc; France; On-site; EUR 34,000 - 42,000; Full time

Job summary

CNRS, Institut des géosciences de l'environnement (IGE) à Grenoble, recherche un postdoc pour modéliser la rheologie anisotrope de la glace et intégrer un réseau neuronal dans le code Elmer/Ice. Le candidat doit détenir un doctorat et démontrer des compétences en calcul numérique et apprentissage automatique.

Le projet est interdisciplinaire et nécessite une collaboration avec plusieurs laboratoires et universités, avec un financement temporaire et un démarrage prévu début 2027.

Qualifications

• Docteur en Sciences avec compétences en calcul numérique et ML.
• Solides bases en méthodes d’homogénéisation et mécanique des microstructures.
• Expérience en rédaction scientifique et travail en équipe.

Responsibilities

• Décrire les microstructures naturelles des glaces avec peu de paramètres.
• Construire une base de données mécanique et évolution microstructure.
• Former et entraîner un réseau neuronal pour prédire la rheologie anisotrope.
• Intégrer le modèle anisotrope dans Elmer/Ice.
• Étudier l’influence de la rheologie sur le flux glaciaire.
• Évaluer les mécanismes de recristallisation et leur impact.

Skills

Numéricale computation

Machine Learning

Approches basées sur les données

Mécanique des matériaux

Education

Ph.D. dans un domaine pertinent

Tools

Elmer/Ice

Job description

Organisation/Company CNRS Department Institut des géosciences de l'environnement Research Field Engineering Physics » Acoustics Engineering » Materials engineering Researcher Profile First Stage Researcher (R1) Application Deadline 25 Nov 2026 - 23:59 (UTC) Country France Type of Contract Temporary Job Status Full-time Hours Per Week 35 Offer Starting Date 1 Jan 2027 Is the job funded through the EU Research Framework Programme? Not funded by a EU programme Is the Job related to staff position within a Research Infrastructure? No

Offer Description

-------Ice flow in polar ice caps and mountain glaciers is largely due to plastic deformation, which can be extremely slow. The rheology of polar ice exhibits strong viscoplastic anisotropy, linked to its microstructure (preferred crystal orientation), which evolves with plastic deformation. Such strongly anisotropic behavior should have a marked influence on the flow dynamics of ice caps, particularly their response to climate transitions, but it has never been fully accounted for in models. The relationship between rheology and microstructure is very well captured by mean-field homogenization methods, which account for ice deformation mechanisms at the fine scale (micrometers) to estimate the effective behavior at the decimeter scale, depending on the material's microstructure. These methods, although very powerful, remain far too computationally expensive to be implemented as a behavior law in a finite element code for polar ice cap flow (scale: 1,000 km).

We therefore propose to train a neural network on a nonlinear homogenization model in order to reproduce the nonlinear and evolving anisotropic rheology of ice, and then to incorporate this neural network into a large-scale flow model. Our initial work in this field has demonstrated the accuracy and extraordinary computational efficiency of AI trained on the homogenization model. We will thus be able to study the impact of ice anisotropy on large-scale flow dynamics and on the evolution of ice sheets in relation to global warming (e.g., as applied to Greenland). This is therefore a highly multidisciplinary numerical project with potential implications for a current societal issue. The researchers involved in this project are experts in the various areas necessary for the successful conduct of the work: mechanical behavior and microstructures of polar ice (Maurine Montagnat, IGE, Olivier Castelnau, PIMM), homogenization methods (Olivier Castelnau, PIMM; Pedro Ponte Castañeda, University of Pennsylvania), numerical calculations of anisotropic geophysical flows (Olivier Gagliardini, IGE; Anne Mangeney, IPGP; Fabien Gillet-Chaulet, IGE), computational mechanics, model reduction and machine learning (Paco Chinesta, Mikhael Tannous, PIMM, Chady Ghnatios, University of North Florida).

Specifically, the postdoctoral work will consist of:

• Describing the various microstructures of natural ice using a small number of parameters, e.g., using generalized spherical harmonics and/or model reduction
• Developing a synthetic database of ice's mechanical response and the evolution of its microstructure, using a recent nonlinear homogenization method
• Develop and train a neural network on this database, capable of characterizing microstructural evolution and predicting anisotropic ice flow
• Implement the anisotropic behavior model based on the neural network into the Elmer/Ice finite element code. This open-source code, used at the Institute of Environmental Geosciences (IGE)—notably by O. Gagliardini's team—enables the simulation of polar ice sheet flow
• Study the influence of incorporating general anisotropic rheology on glacial flow dynamics and the evolution of polar ice caps
• Extend and apply the developed model to various ice recrystallization mechanisms in order to evaluate their influence on microstructural evolution and on the anisotropic mechanical behavior of ice

The Institute of Environmental Geosciences (IGE) is a public research laboratory affiliated with the CNRS, the IRD, the University of Grenoble Alpes (UGA), and Grenoble-INP. Its research focuses on climate change and the impact of human activities on our planet in polar, mountainous, and intertropical regions—areas that are particularly sensitive to major societal challenges. The laboratory has an average staff of approximately 330 people, including 190 permanent employees (researchers, faculty researchers, engineers, technicians, and administrative staff) and approximately 140 doctoral students, postdoctoral researchers, and contract employees. Each year, the IGE also hosts approximately 120 interns and visiting scientists. The IGE is spread across four buildings on the Grenoble university campus (the Glaciology Building, OSUG-B, the House of Climate and the Planet, and INRAE-Grenoble in Saint-Martin-d'Hères). This work is part of a program funded by the Ministry of Ecological Transition (PAPROG) that aims to improve the Elmer/Ice glacier model in order to better predict glacier-related risks. The postdoctoral researcher will be based at the Institute of Environmental Geosciences (IGE, Grenoble), where he or she will work with Olivier Gagliardini, Maurine Montagnat, and Fabien Gillet-Chaulet in close collaboration with Olivier Castelnau, Francisco Chinesta and Mikhael Tannous from the PIMM laboratory, Anne Mangeney from the Institute of Earth Physics (IPGP, Paris), as well as Chady Ghnatios (University of North Florida, United States) and Pedro Ponte Castañeda (University of Pennsylvania, United States).

The candidate, who must have earned a Ph.D., must have demonstrated strong skills in numerical computation, machine learning, data-driven approaches, and materials or structural mechanics. Knowledge of homogenization methods, the microstructures of polycrystalline materials, and geophysics will also be appreciated. The candidate must have the ability to write scientific articles and be motivated to work as part of a team on a topic at the intersection of several disciplines.