Research Engineer for climate data downscaling (M/F)

Job Description

Missions

The mission for this position is to extend and operationalize the IPSL-AID generative downscaling tool for coastal ocean applications. The engineer will design and integrate a specialized processing pipeline into the existing model architecture to handle the acquisition, preprocessing, and specific conditioning of coastal data (e.g., high-resolution reanalysis and satellite observations). He/she will also adapt the model's inference process to produce very high-resolution (~2-4 km) probabilistic projections with uncertainty quantification over target coastal zones.

Activity

The main activities associated with this position include:
● Analyzing the existing IPSL-AID model architecture to identify optimal extension points for the integration of the new processing pipeline.
● Designing and integrating a new pipeline for the automated preprocessing of specific input datasets (MERCATOR ocean reanalyses and satellite observations), including data formatting for the model and their incorporation into the training workflow.
● Developing, implementing, and validating novel conditioning schemes (e.g., via embeddings or concatenation) enabling IPSL-AID to effectively incorporate sparse input data.
● Adapting the model's inference and sampling scripts to handle regional coastal domains, including boundary treatment and tile stitching.
● Establishing a validation protocol for coastal outputs, leveraging satellite and in-situ observations to assess model performance.
● Packaging and deploying the integrated coastal pipeline as a Python library, with clear API documentation and usage examples, for distribution to project partners.

Your Profil

Skills

Education
PhD or Master's degree (Bac +5) in Computer Science, Data Science, Software Engineering, Physical Oceanography/Climate Science, or a closely related field.
Required Technical Skills
● Production-Grade Python Development: Advanced proficiency in Python, with strong experience in developing APIs, modular libraries, and reproducible data pipelines. Familiarity with the scientific Python ecosystem (xarray, pandas, NumPy, etc.) is required.
● Machine Learning / Deep Learning Models: Hands-on experience with PyTorch (required) for loading, running, and extending existing models. A solid understanding of U-Net–type architectures and conditioning mechanisms is expected. Prior exposure to diffusion or other generative models is a strong asset.
● Data Engineering: Strong skills in NetCDF data handling, with experience in processing climate and/or oceanographic datasets being highly desirable.
● Computing Environments: Demonstrated experience with HPC environments (job scheduling, resource management) and GPU acceleration.
Major Assets
● Experience integrating or extending large-scale research models (in climate science, meteorology, or related domains).
● Experience with modern Python package management tools (e.g., pixi, uv).
● Knowledge of computational server infrastructures.
Personal Qualities
● Strong software engineering skills, with a high standard for code quality, testing, and documentation.
● Ability to work autonomously on complex technical development tasks, combined with excellent collaboration skills within a research team.
● A pragmatic, solution-oriented approach to integrating new functionalities into an existing codebase.
● Rigor in result validation and experimental reproducibility.
● Ability to work closely with other IPSL-AID developers and to interact with end users (e.g., ecologists, oceanographers) to gather requirements and validate outputs.
● Strong written and oral communication skills.

Your Work Environment

Le poste est basé au Centre de Modélisation Climatique de l'Institut Pierre-Simon Laplace (CMC-IPSL), situé au campus Pierre et Marie Curie de l'Université Sorbonne, 75005 Paris. IPSL est un institut de renommée mondiale, reconnu pour ses recherches de pointe en sciences du climat, de l'environnement et des systèmes terrestres à travers la modélisation.
Les conditions de travail incluent une possibilité de télétravail partiel, un accès à des équipements informatiques de haute performance, et des formations professionnelles CNRS. Le salaire est basé sur l'expérience du candidat et la grille salariale du CNRS. Des avantages sociaux du CNRS, incluant l'assurance santé, les plans de retraite, ainsi que des subventions pour le transport public et les repas, sont également offerts.

Compensation and benefits

Compensation

Starting at €3,237.95 gross per month, depending on experience

Annual leave and RTT

44 jours

Remote Working practice and compensation

Pratique et indemnisation du TT

Transport

Prise en charge à 75% du coût et forfait mobilité durable jusqu’à 300€

About the offer

About the CNRS

The CNRS is a major player in fundamental research on a global scale. The CNRS is the only French organization active in all scientific fields. Its unique position as a multi-specialist allows it to bring together different disciplines to address the most important challenges of the contemporary world, in connection with the actors of change.

CNRS

The research professions