Cerfacs Enter the world of high performance ...

The 13 December 2021 at 10h00

PhD Defense: Camille BESOMBES : Parameterization using GANs for reduction of control space in data assimilation

Nathalie BROUSSET |  CERFACS, Toulouse, Jean-Claude ANDRÉ Meeting Room  |  

Abstract :

This thesis examines the use of generative adversarial networks (GANs) as a parameterization tool for inverse problems solved with ensemble-based data assimilation methods. Ensemble methods often rely on the assumption of Gaussian distributed parameters in cases where this assumption is not valid, the parameter estimation can be invalid. Parameterization methods allow the transformation of these non-Gaussian parameters into a better suited distribution, and optimally reduce their dimension.
Another limitation of ensemble methods is the injection of prior information of the physical relation as a constraint between parameters such as spatial coherence or physical balances. Optimal parameterization should encompass these different properties to facilitate the estimation. The novel approach presented in this work relies on GANs to achieve these objectives. Two application domains are tackled through the present work.

In a first application, subsurface reservoir characterization, the objective is to determine geological properties of a numerical reservoir model from the observation of the reservoir dynamical response by the way of data assimilation. Rock facies, that describe the type of rock present in each cell of the numerical model, have to be determined due to their strong influence on the dynamical response. The rock facies spatial distribution is ruled by geological phenomena such as sedimentation and forms well known patterns, like channels, called heterogeneities. The non-continuous property and their spatial coherence make their characterization by ensemble-based data assimilation algorithms difficult, and requires parameterization. Parameterization is a challenge for numerous heterogeneities, notably channels, due to the numerical cost or the statistical representation of their spatial distribution.

A Second application domain is the atmospheric balance in the context of numerical weather prediction. When new observations are available, correction of the atmospheric state is done using ensemble based data assimilation methods. This correction step can introduce imbalance in the physical state and cause numerical instability during the integration in time of the atmosphere, deteriorating the information brought by the previous observations. The importance of generating or correcting balanced climate, also called initialized atmospheric state, during data assimilation is then a key step in numerical weather prediction.

This work aims at presenting the performance of GAN parameterization and its multi-disciplinary applicability to researchers who are not familiar with the domain of deep learning. GAN is an unsupervised deep learning method belonging to the deep generative network family, able to learn a dataset distribution and generate new samples from the learned distribution in an unsupervised way. These synthetic samples are encoded in a low-dimensional latent space that can be sampled from a Gaussian distribution that is suited to perform ensemble data assimilation. Their recent ability to generate complex images led us to consider them as a good candidate for parameterization method. The unsupervised property of this type of parameterization makes it applicable to several diverse domains such as learning the pattern of geological heterogeneities or learning the physical constraints that makes an atmospheric state balanced.

This study shows how to train GANs for two different applications : subsurface reservoir and climate data. The use of the parameterization in an ensemble based data assimilation such as ensemble smoother with multiple data assimilation (ES-MDA) is demonstrated for subsurface reservoirs. Finally, a posteriori conditioning of the GAN function is examined using derivative free optimization.

Jury :

Olivier THUALINPAdvisor
Eric BLAYOUniversité Grenoble AlpesReferee
Ronan FLABETIMT AtlantiqueReferee

Corentin LAPEYRE


Toulouse INP/IMFT
















Météo France – CNRM











Invited member

nvited member












Sparse Days in Saint-Girons IV

Brigitte Yzel |  29 March 2022

Sparse Days Meeting 2022 🗓️  20-22 June 2022  @ Saint-Girons, France   🌐   Sparse Days 2022 will be held in Saint-Girons, Ariège, from 20-22 June. This enhanced version of Sparse Days is being co-organized by Cerfacs and ENSEEIHT/IRIT. It will be the fourth meeting in Saint- Girons following the tradition of the previous meetings held in 1994, 2003, and 2015. The tradition involves coupling our highly successful annual technical meeting with the ambience and hospitality of this wonderful Pyrenean town which encourages fruitful informal exchanges between participants.    Read more

Press releases on the NextSim EuroHPC project

CERFACS |  9 February 2022

Press talks about the NextSim EuroHPC project. CERFACS is involved in two major points. The first topic is to extend the numerical capabilities of the code to perform scale-resolving simulations of jet noise configurations on exascale supercomputers. The second focus is to exploit data from these simulations by extracting or visualising relevant physical information during the computation. CERFACS is involved in the NextSim project (). The primary objective is to increase the capabilities of Computational Fluid Dynamics tools on extreme-scale parallel computing platforms for aeronautical design. This project has received funding from the European High-Performance Computing Joint Undertaking (JU) under grant agreement N° 956104. The JU receives support from the European Union’s Horizon 2020 research and innovation programme and Spain, France, Germany. This project has received funding from the Agence Nationale de la Recherche (ANR) under grant agreement N° ANR-20-EHPC-0002-02. For more information, please visit Read more