Come join IMT Mines Alès, a prestigious "grande école" ranked among the top engineering schools both nationally and globally, located in Alès---a welcoming, livable city and the capital of the Cévennes region, where residents highly value the quality of life.

Founded in 1843, IMT Mines Alès currently has 1,400 students (including 250 international students) and 380 staff members. The school has three high-level scientific and technological research and teaching centers, working in the fields of materials and civil engineering (C2MA), the environment and risk management (CREER), and artificial intelligence and industrial and digital engineering (CERIS). It operates 12 technology platforms and has 1,600 partner companies. For more information: https://www.imt-mines-ales.fr

You will be affiliated with the **Center for Research and Teaching in Environment and Risks (CREER)** at IMT Mines Alès and the **HSM** research unit.

This work is part of the **PREV'IA** project, which focuses on understanding and forecasting water resources in the Pliocene-Quaternary aquifer of the Roussillon Plain and the Lez spring that supplies Montpellier with drinking water.

Your work will build upon an ongoing thesis within the UMR HSM (HYTAKE team), focusing on the sustainability of the karst water resource (Lez aquifer) supplying a Mediterranean metropolis (Montpellier) in the context of climate change.

**Your mission:**

* Use emerging concepts in "Physics-Informed Machine Learning" (PIML) approaches to assess the potential benefits of synergy in hydrogeological modeling: (1) concept-based modeling on the one hand, and (2) neural networks informed by, or complemented with, physics on the other. The PIML (Physics Informed Machine Learning) approach is being developed along two lines: (i) Physics Informed Architecture (PIA): integration of physical knowledge into the model architecture (hidden layers); and (ii) Physics Informed Loss Function (PILF): combining physical modeling (differential equations) and black-box methods, merged into the cost function (regularization term). Both methods may be explored. The proposed models may be used to determine the evolution of certain hydrological variables in response to climate change.

You can find more information in the **job description** below:

https://partage.imt.fr/index.php/s/mDKCaJkyYTMmepM