Cerfacs Enter the world of high performance ...

Data assimilation

  From Monday 17 June 2019 to Wednesday 19 June 2019


Deadline for registration: 15 days before the starting date of each training
Duration : 3 days / (21 hours)




Data assimilation is an important component of modelling in a number of applications in the geosciences and in engineering. This training course will provide an overview of the theory and practice of data assimilation. First, the basic concepts from statistical estimation theory and nonlinear optimization will be given. The classical variational and Kalman filtering approaches to data assimilation will then be described. The lectures will also touch upon more specialized topics including covariance modelling and estimation, advanced minimization algorithms, preconditioning, and hybrid ensemble-variational methods. The lectures on the theory will be complemented by both practical exercises and presentations on specific applications at CERFACS in the geosciences (oceanography, atmospheric chemistry, and/or hydrology/hydraulics).

Objective of the training

The objectives are to learn the key aspects of data assimilation, to understand the theory behind the methodologies, and to make the link with state estimation in the geosciences through specific applications.

Learning outcomes

On completion of this training you will be able to:

  • give a grounded opinion about advanced methods in data assimilation: variational assimilation, the Kalman filter, and their ensemble variants;
  • apply data assimilation and covariance modelling methods to simple problems;
  • appreciate the importance of and critical issues in data assimilation for real-world applications.

Target participants

This training session is for engineers, physicists, computer scientists and numerical analysts wishing to learn the fundamentals of data assimilation and the numerical methods to develop data assimilation applications.


In order to follow this course, you need to:

  • know basics methods in numerical linear algebra,
  • have a background in statistics and probability theory.

To verify that the prerequisites are satisfied, the following questionnaires must be completed. You need to get at least 75% of correct answers in order to be authorized to follow this training session. If you don’t succeed it, your subscription will not be validated. You only have two chances to complete them.

Questionnaire 1 : Numerical Analysis

Questionnaire 2 : Statistical Analysis

Scientific contacts: Selime Gürol & Anthony Weaver


  • Trainees/PhDs/PostDocs :  210 €
  • CERFACS shareholders/CNRS/INRIA : 600 €
  • Public : 1200 €


(Every day from 9h to 17h30)

Day 1:

Basic concepts and methodologies of data assimilation
Application from Earth sciences

Day 2:

Variational data assimilation
Covariance modelling and estimation

Day 3:

The Kalman filter and Ensemble Kalman filter
Hybrid ensemble-variational methods
Application from Earth sciences

Final examination

A final exam will be conducted during the training.






Multi-architecture parallelism using Kokkos/C++ library

From Monday 11 December 2023 at 14h00 to Wednesday 13 December 2023 at 17h00






PhD Defense : Aurélien LINÉ : ” Modulation of European near-term climate change by multi-decadal internal variability “

Thursday 21 December 2023 at 15h00

  CERFACS - Toulouse - France     Organized by Nathalie BROUSSET    





Introduction to programming practices for scientific computing

Monday 18 March 2024