PIF2025 - Deep learning applied to ocean modelling - (PID2024-159029NB-C21)

The selected candidate will conduct cutting-edge research to bridge the gap between Physical Oceanography and Artificial Intelligence. The core objective of the PhD is to expand novel hybrid downscaling methods, specifically developing statistical Ocean Model (OM) emulators. The student will adapt and apply advanced Deep Learning to reproduce complex ocean variables. A critical part of the work involves applying eXplainable Artificial Intelligence (XAI) techniques to unravel the "black box" nature of these models, ensuring their physical consistency and robustness for climate projections. Training & Mobility: Designed for high-level training, the plan includes: • International stays • Participation in summer schools and workshops. • Publication in high-impact SCI journals. Requirements • Master’s in Physics, Maths, Engineering, Data Science, or Marine Sciences. • Background in Physical Oceanography OR in Machine Learning/Deep Learning. Candidates from Physics/Maths willing to learn oceanography, or Oceanographers with strong coding skills, are encouraged to apply. • Skills: Proficiency in Python or MATLAB. Experience with ocean models or climate data is a plus. What We Offer • Environment: You will join a small, highly motivated research team known for its dynamic atmosphere and strong international connections. The position integration into the PhD program in Physics at UIB. • Training Support: Access t