Cerfacs Enter the world of high performance ...

Climate Extreme Events Analysis using Machine Learning Methods

  |   |  , , ,

Required Education : Engineer or Master 2
Start date : 2 January 2023
Mission duration : 12 months (extensible to 18 months)
Deadline for applications : 22 December 2022
Salary : 2300 eur/month (according to experience)


In the context of a Horizon Europe project on Digital Twins and DestinE, interTwin https://www.intertwin.eu/ , the analysis of a great number of climate simulations of the future climate is necessary in order to evaluate the change of characteristics and the occurrence of extreme events. Traditional analysis techniques do not allow an efficient and reasonable performance, and neither to benefit from the new GPU computing architectures. In the context of Digital Twins, it is essential to develop novel analysis methods using artificial intelligence techniques. This work is being done in constant collaboration with European and international partners.


This position requires solid knowledge of statistics and probability theory, given that the principal objective is to develop a novel method to characterize extreme events in the climate simulations using methods based on artificial intelligence (Machine Learning). This objective aims at assessing the change of characteristics of extreme meteorological events in the future climate: occurrence frequency, spatial extent, temporal extent, intensity. Input datasets are reanalysis and global climate simulations for the historical and future climate period of CMIP6.

Knowledge of programming and development, ideally in python, is necessary, as well as the knowledge about scientific data processing. It is also an advantage to know how to apply Machine Learning methods (also Deep Learning and Data Mining), and knowledge about the following is also considered: Linux environment, github/gitlab, etc.

Work will be done in collaboration with several French and European partners, according to specifications written in the Description of Work of the project, as well as the result of discussions during European meetings and workshops. This position implies active participation and the presentation of work achieved in numerous visio-conferences, meetings and workshops. It also includes deliverable writing (in english) for the European Commission in the context of the project, explaining the work achieved. The project has 26 partners.

The work premises is at CERFACS, in a multi-disciplinary and dynamic team, with PhD students, post-docs, researchers and research engineer. The computer hardware is very recent, with a very fast and flexible IT team support, with access to internal and external supercomputer resources. The candidate must be at ease and efficient when working within a team, fluent in english, and serious and dedicated to work combined with an excellent autonomy. Remote working is also possible, up to 2 days per week. It is important to mention that work autonomy is crucial to succeed in this position.


Christian Pagé christian.page@cerfacs.fr