POSTDOCTORAL RESEARCHER (M/F) – MACHINE‑LEARNING SPECIALIST FOR DOWNSCALING VIA REGIONAL CLIMATE‑MOD
Contrat : CDD Temps Plein
Societe : Météo France
Localisation : 31057 Toulouse | (FR)
Description du poste :
As a public expert in weather and climate, Météo-France is here to help keep you safe every day and assist you in making the best decisions in a changing climate. With dangerous weather events becoming even more intense and frequent due to climate change, our mission to keep you safe is crucial. We mobilize our expertise and scientific and technological excellence to enable you to anticipate and adapt to challenging weather and climate events.
Find us online:https://meteofrance.com/carte-didentite-de-meteo-france
The recruitment is part of the TRACCS (TRAnsforming Climate modelling for Climate Services, https://pepr-traccs.fr/) research programme, which brings together the French climate modelling community. The programme's activities cover the fundamental understanding of climate change and its impacts, and extend to the development of prototype climate services co-constructed by stakeholders and climate modelling experts. The aim is to accelerate the development of climate models to meet society's expectations in terms of climate action, particularly in the area of adaptation to climate change.
The programme is organised into 10 targeted projects and a governance project, and will be supplemented by projects in response to calls for tender. It has been allocated €51 million over 10 years. It is co-piloted by CNRS and Météo-France, with 7 other academic partners. The activities of the governance project and the targeted projects will be carried out mainly in the Paris region (laboratories of the Institut Pierre-Simon Laplace (IPSL)), in Toulouse (CNRM and other Météo-France entities, CERFACS) and in Grenoble (Institut des Géosciences de l'Environnement (IGE)).
The present position is part of the TRACCS-PC10-LOCALISING (https://pepr-traccs.fr/projet/pc10-localising/) project, which aims to explore and define how best to provide accurate and reliable local climate information in support of adaptation strategies. To achieve this, LOCALISING is developing fully coupled, multi-component models of local climate systems, aiming to represent the climate at kilometre and hour scales, and combining dynamic models and statistical approaches to characterise uncertainty at local scales. A number of technical and scientific challenges will have to be overcome in order to achieve the project's ambitious objectives:
(1) increase the resolution of climate models and resolve the resulting bottlenecks,
(2) develop models to represent new local processes and feedbacks at these higher resolutions,
(3) study the key physical, chemical and biogeochemical processes at the local scale in order to gain a better understanding of climate phenomena, in particular the extremes that are relevant to society,
(4) take advantage of emerging machine learning techniques to quantify uncertainty at lower cost, by developing new statistical and hybrid downscaling techniques,
(5) ensure the consistency of climate information throughout the modelling chain, from global climate models to fine-scale climate data,
(6) ensure the transportability of the methods developed throughout the world and support a wide range of users.
The DESR brings together the research entities of Météo-France (mainly CNRM, SAFIRE and LACy), the National School of Meteorology (ENM) and the shared administrative and IT support services (PGA).
The CNRM is a Joint Research Unit (UMR 3589, www.umr-cnrm.fr) under the joint supervision of Météo-France and the CNRS. The CNRM conducts research in the field of meteorology and climate, from the observation, understanding and modelling of processes to the development of weather forecasting and climate projection systems that can be transferred to Météo-France's operational services
The person recruited will join the COMETS team (Chemistry: Observation and Modelling of thE Troposphere and Stratosphere, https://www.umr-cnrm.fr/spip.php?article371&lang=en) of the CNRM's Groupe de Modélisation Grande
The newly hired researcher will join the MOSCA team (Modélisation du Système Climatique Régional, https://cnrm.sedoo.fr/mosca‑modelisation‑du‑systeme‑climatique‑regional/) within the Large‑Scale Modeling and Climate Group (GMGEC) at CNRM. The team’s goal is to improve knowledge of regional climates through a dedicated modeling approach.
Why join us?
Embark on a stimulating adventure that serves everyone, alongside men and women who are committed every day to tackling the challenges posed to our society by weather and climate.
And enjoy the following benefits: flexible working hours, RTT (reduced working time), teleworking, staff restaurant or meal vouchers, 75% contribution to public transport costs, contribution to mutual insurance, sports and cultural associations depending on the site concerned (climbing, gym, pottery, theater, etc.).
Other benefits await you, come and discover them!
Context
In recent years, thanks to the development of modern statistical‑learning methods, emulators of regional climate models (RCMs) have emerged as a promising solution for providing regional climate simulations allowing to reliably studying the impacts of climate change at local scales. The main objective is to learn a downscaling relationship between a “large‑scale” description of the atmospheric situation (low‑resolution, mid‑to‑upper‑troposphere) and one or several high‑resolution surface variables (those produced by the RCM). Once estimated, this downscaling function can be applied to any simulation from a low‑resolution global model in order to generate high‑resolution simulations. Very large ensembles can thus be created to study local climate change while accounting for the multiple sources of uncertainty inherent to climate projections. A first version of an RCM emulator has been developed and used at CNRM (Doury et al., 2023, 2024). It is an emulator of the ALADIN regional climate model for daily temperature and precipitation at a 12 km resolution.
Objectives
Within the LOCALISING project of the TRACCS program, the development of the CNRM emulator will be continued. Indeed, many scientific and technical hurdles still need to be overcome in order to obtain a tool that retains as much as possible the characteristics of a high‑resolution climate model while keeping computational costs to a minimum. RCM emulators open the door to new possibilities for evaluating the impacts of climate change locally thanks to the very large ensembles that can be produced. The analysis of these ensembles therefore raises new questions.
The position described here fits within this framework; the selected candidate will continue a research program aimed at advancing knowledge on the development of these tools so as to better characterize the local effects of climate change. Candidates are asked to submit a research project of no more than two pages to justify and outline the lines of inquiry they wish to pursue. The successful applicant will take part in the various project activities (meetings, workshops, conferences, deliverables, etc.) and will develop collaborations with the other project partners (LSCE, CECI). Dissemination of the research results through scientific publications or conference presentations is also expected.
Application deadline : 30th April 2026
For any further information : antoine.doury@meteo.fr / samuel.somot@meteo.fr
General Knowledge
The successful candidate must hold a Ph.D. in one of the following fields: applied mathematics, environmental sciences, or computer science.
Experience in machine learning is required, as well as a good understanding of basic statistical concepts, and familiarity with neural‑network frameworks is a plus. Prior experience with climate studies, climate change, and/or climate models is also desirable.
Technical Skills
* Excellent proficiency in Python.
* Mastery of neural‑network development tools (e.g., Keras, PyTorch).
* Proficiency with version‑control systems (e.g., Git).
* Strong command of the Unix environment.
* Experience handling NetCDF files (e.g., NCO, CDO).
Core Competencies
* Ability to work autonomously and organise one’s tasks efficiently.
* Rigorous approach to software development and scientific analysis.
* High level of motivation and scientific curiosity.
* Oral and written English at least at B2 level.
Personal Qualities
* Good interpersonal skills and a strong sense of teamwork: you will work in a collaborative setting, contributing to a large, multidisciplinary team.
* Networking ability: you will be part of the TRACCS community and will help foster its development and dynamism.
* Availability and responsiveness.
* Professional kindness and a supportive attitude toward colleagues.
Salaire : Non communiqué
Profil :
Etude / Diplôme souhaité :
Expérience souhaitée : 2 à 4 ans
Réf. : 7816760
Offre d'emploi publiée le 2026-04-17 2026-06-01