The article of Camille Besombes (PhD student)
"Producing realistic climate data with generative adversarial networks" is now published in NPG 10.5194/npg-28-347-2021

This research article shows the ability of Wasserstein GAN to generate realistic 3D global weather situation, This opens the way to new data assimilation techniques for weather prediction and risk management.

Summary

This paper investigates the potential of a Wasserstein generative adversarial network to produce realistic weather situations when trained from the climate of a general circulation model (GCM). To do so, a convolutional neural network architecture is proposed for the generator and trained on a synthetic climate database, computed using a simple three dimensional climate model: PLASIM.
The generator transforms a “latent space”, defined by a 64-dimensional Gaussian distribution, into spatially defined anomalies on the same output grid as PLASIM. The analysis of the statistics in the leading empirical orthogonal functions shows that the generator is able to reproduce many aspects of the multivariate distribution of the synthetic climate. Moreover, generated states reproduce the leading geostrophic balance present in the atmosphere.
The ability to represent the climate state in a compact, dense and potentially nonlinear latent space opens new perspectives in the analysis and handling of the climate. This contribution discusses the exploration of the extremes close to a given state and how to connect two realistic weather situations with this approach.

The article
"An anisotropic formulation of the parametric Kalman filter assimilation" is now published in Tellus A 10.1080/16000870.2021.1926660

This research article detail how to assimilate local observations and how predict the dynamics of the error covariance matrix with the PKF. The work relies on the parametric Kalman filter introduced in two
previous contributions
(Pannekoucke et al. 2016,
Pannekoucke et al. 2018,
)

Summary

In geophysics, the direct application of covariance matrix dynamics described by the Kalman filter (KF) is limited by the high dimension of such problems. The parametric Kalman filter (PKF) is a recent alternative to the ensemble Kalman filter, where the covariance matrices are approximated by a covariance model featured by a set of parameters. The covariance dynamics is then described by the time evolution of these parameters during the analysis and forecast cycles. This study focuses on covariance model parametrized by the variance and the local anisotropic tensor fields (VLATcov). The analysis step of the PKF for VLATcov in a 2D/3D domain is first introduced. Then, using 2D univariate numerical investigations, the PKF is shown to be able to provide a low numerical cost approximation of the Kalman filter analysis step, even for anisotropic error correlation functions. Moreover the PKF has been shown able to reproduce the KF over several assimilation cycles in a transport dynamics. An extension toward the multivariate situation is theoretically studied in a 1D domain.

The article "A methodology to obtain model-error covariances due to
the discretization scheme from the parametric Kalman filter perspective"
is now published in Nonlin. Processes Geophys
https://doi.org/10.5194/npg-28-1-2021

This research article estimate for the first time the model-error
covariance statistics due to the discretization of the partial differential
equation. The work relies on the parametric Kalman filter introduced in two
previous contributions
(Pannekoucke et al. 2016,
Pannekoucke et al. 2018,
)

Summary

This contribution addresses the characterization of the model-error covariance
matrix from the new theoretical perspective provided by the parametric Kalman filter
method which approximates the covariance dynamics from the parametric evolution of
a covariance model. The classical approach to obtain the modified equation of
a dynamics is revisited to formulate a parametric modelling of the model-error
covariance matrix which applies when the numerical model is dissipative compared
with the true dynamics. As an illustration, the particular case of the advection
equation is considered as a simple test bed. After the theoretical derivation
of the predictability-error covariance matrices of both the nature and the
numerical model, a numerical simulation is proposed which illustrates the
properties of the resulting model-error covariance matrix.

The article "PDE-NetGen 1.0: from symbolic PDE representations of physical processes to trainable
neural network representations" is now published in Geo. Mod. Dev.
https://doi.org/10.5194/gmd-13-3373-2020

This research article bridges physics and design of neural network.
The ongoing version of the code is on github
https://github.com/opannekoucke/pdenetgen
(a snapshot is available here:
)

Summary

Bridging physics and deep learning is a topical challenge. While deep learning frameworks open avenues
in physical science, the design of physicallyconsistent deep neural network architectures is an open issue.
In the spirit of physics-informed NNs, PDE-NetGen package provides new means to automatically translate
physical equations, given as PDEs, into neural network architectures. PDE-NetGen combines symbolic calculus
and a neural network generator. The later exploits NN-based implementations of PDE solvers using Keras.
With some knowledge of a problem, PDE-NetGen is a plug-and-play tool to generate physics-informed NN
architectures. They provide computationally-efficient yet compact representations to address a variety of
issues, including among others adjoint derivation, model calibration, forecasting, data assimilation as
well as uncertainty quantification. As an illustration, the workflow is first presented for the 2D diffusion
equation, then applied to the data-driven and physics-informed identification of uncertainty dynamics for the
Burgers equation.

Invited talk at the AI4OAC
workshop, Brest, 20-25 January, 2020

In this talk I'll present neural network generator for evolution equations.

Talk at the Climath WK2
"Big Data, Data Assimilation, Uncertainty quantification",
Institute Henri Poincaré, Paris, 12-15 November, France, 2019.

In this talk I'll present symbolic tools for the computation of the parametric Kalman filter
dynamics and the neural network generator for evolution equations. In particular, I'll show
how to merge known physical dynamics and deep learning to learn unknown processes.

This research article shows the ability of the parametric Kalman filter to predict the dynamics of the uncertainty
at the tangent-linear approximation in the Burgers equation, and without any ensemble estimation.

Short Summary

The forecast of weather prediction uncertainty is a real
challenge and is crucial for risk management. However, uncertainty prediction is
beyond the capacity of supercomputers, and improvements of the technology may not
solve this issue. A new uncertainty prediction method is introduced which takes advantage of
fluid equations to predict simple quantities which approximate real uncertainty but at a low
numerical cost. A proof of concept is shown by an academic model derived from fluid dynamics.

Curiculum Vitae

Education

Positions

research interests

INPT ENM
42, Av Gaspard Coriolis
31057 Toulouse cedex 1
France

A. Perrot, O. Pannekoucke and V. Guidard, Toward a multivariate formulation of the PKF ssimilation: application to a simplified chemical transport model. submitted egusphere-2022-928

M. Sabathier, O. Pannekoucke, V. Maget, and N. Dahmen, Boundary Conditions for the Parametric Kalman Filter forecast submited, 2022. submitted
doi: 10.1002/essoar.10512724.1

Published articles

R. Fablet, B. Chapron, L. Drumetz, E. Mémin, O. Pannekoucke and F. Rousseau,
"Learning Variational Data Assimilation Models and Solvers" Journal of Advances in Modeling Earth Systems, vol. 13, no. 10, p. e2021MS002572, 2021
https://doi.org/10.1029/2021MS002572, 2021.

O. Pannekoucke and P. Arbogast,
“SymPKF (v1.0): a symbolic and computational toolbox for the design of parametric Kalman filter dynamics,” Geosci. Model Dev., 14, 5957–5976, 2021
doi: 10.5194/gmd-14-5957-2021

R. Ménard, S. Skachko, and O. Pannekoucke,
“Numerical discretization causing error variance loss and the need for inflation,” Quarterly Journal of the Royal Meteorological Society, Aug. 2021,
doi: 10.1002/qj.4139

C. Besombes, O. Pannekoucke, C. Lapeyre, B. Sanderson, and O. Thual,
“Producing realistic climate data with generative adversarial networks” Nonlinear Processes in Geophysics, vol. 28, no. 3, pp. 347–370, Jul. 2021,
doi: 10.5194/npg-28-347-2021

O. Pannekoucke,
An anisotropic formulation of the parametric Kalman filter assimilation,Tellus A: Dynamic Meteorology and Oceanography, vol. 73, no. 1, pp. 1–27, Jan. 2021, doi: 10.1080/16000870.2021.1926660.

O. Pannekoucke and R. Fablet,
PDE-NetGen 1.0: from symbolic PDE representations of physical processes to trainable
neural network representations, Geosci. Model Dev., 13, 3373–3382, 2020.
https://doi.org/10.5194/gmd-13-3373-2020,
snapshot of the code associated to the article:
ongoing version on github: https://github.com/opannekoucke/pdenetgen

O. Pannekoucke Ricci, S.; Barthelemy, S.; Ménard, R. & Thual., O.
Parametric Kalman filter for chemical transport models - Corrigendum
Tellus A: Dynamic Meteorology and Oceanography, 70, 1-2 (2018)
link

O. Pannekoucke P. Cebron, N. Oger, and P. Arbogast.
From the Kalman Filter to the Particle Filter: A geometrical perspective of
the curse of dimensionality.
Advances in Meteorology, 2016, 9372786 (2016)
link

O. Pannekoucke, S. Ricci, S. Barthelemy, R. Menard and O. Thual,
Parametric Kalman filter for Chemical Transport Models.
Tellus A, 68:31547, (2016).
https://doi.org/10.3402/tellusa.v68.31547

Ph. Arbogast, O. Pannekoucke, L. Raynaud, R. Lalanne, and E. Memin.
Object-oriented processing of CRM precipitation forecasts by stochastic filtering.
Q. J. R. Meteorol. Soc. 142:2827--2838 (2016)
link

R. Mechri, C. Ottle, O. Pannekoucke, A. Kallel, F. Maignan, D. Courault and I. Trigo,
Downscaling Meteosat Land Surface Temperature over a Heterogeneous Landscape Using a Data Assimilation Approach Remote Sensing,
MDPI AG, 2016, 8, 586. (2016)
link

L. Raynaud, O. Pannekoucke, P. Arbogast, and F. Bouttier.
Application of a Bayesian weighting for short-range lagged ensemble forecasting at convective scale.
Q. J. R. Meteorol. Soc. 141:459--468 (2014)
link

R. Mechri, C. Ottle, O. Pannekoucke and A. Kallel.
Genetic Particle Filter application to Land Surface Temperature downscaling
in Journal of Geophysical Research. (2014)
link

E. Emili, B. Barret, S. Massart, E. Le Flochmoen, A. Piacentini, L. El Amraoui, O. Pannekoucke, and D. Cariolle.
Combined assimilation of IASI and MLS observations to constrain tropospheric and stratospheric ozone in a global
chemical transport model. Atmos. Chem. Phys., 14, 177-198 (2014)
link

M. Zamo, O. Mestre, Ph. Arbogast, and O. Pannekoucke,
A benchmark of statistical regression methods for short-term forecasting of photovoltaic electricity production,
part I: deterministic forecast of hourly production.
Solar Energy (2014)
link

M. Zamo, O. Mestre, Ph. Arbogast, and O. Pannekoucke,
A benchmark of statistical regression methods for short-term forecasting of photovoltaic electricity production,
part II: probabilistic forecast of daily production.
Solar Energy (2014)
link

O. Pannekoucke, E. Emili and O. Thual,
Modeling of local length-scale dynamics and isotropizing deformations,
Q. J. R. Meteorol. Soc. (2014)
link

M. Boisserie, Ph. Arbogast, L. Descamps, O. Pannekoucke, L. Raynaud.
Estimating and diagnosing model error variances in the Meteo-France global NWP model,
Q. J. R. Meteorol. Soc. (2014)
link

O. Pannekoucke, L. Raynaud and M. Farge,
A wavelet-based filtering of ensemble background-error variances,
Q. J. R. Meteorol. Soc. 140:846--854 (2014)
link

L. Raynaud and O. Pannekoucke,
Sampling properties and spatial filtering of ensemble background-error length-scales,
Q. J. R. Meteorol. Soc. 139:784--794 (2013)
link

L. Raynaud and O. Pannekoucke.
Heterogeneous filtering of ensemble-based background-error variances.
Q. J. R. Meteorol. Soc. 138: 1589--1598 (2012)
link

S. Massart, A. Piacentini, and O. Pannekoucke.
How important is to use diagnosed background error covariances for the atmospheric ozone analysis?
Q. J. R. Meteorol. Soc. 138: 889--905 (2012)
link

N. Oger, O. Pannekoucke, A. Doerenbecher and P. Arbogast.
Assessing the trajectory influence in adaptive observation Kalman filter
sensitivity method.
Q. J. R. Meteorol. Soc.
138: 813--825 (2012)
link

S. Remy, O. Pannekoucke, T. Bergot and C. Baehr.
Adaptation of a particle filtering method for data assimilation in a 1D numerical model used for fog forecasting.
Q. J. R. Meteorol. Soc. 138: 536--551 (2012)
link

S. Massart, B. Pajot, A. Piacentini and O. Pannekoucke.
On the merits of using a 3D-FGAT assimilation scheme with an outer loop for atmospheric
situations governed by transport.
Mon. Wea. Rev. 138:4509-4522. (2010)
link

O. Pannekoucke and L. Vezard.
Stochastic integration for the heterogeneous correlation modeling
using a diffusion equation.
Mon. Wea. Rev. 138: 3356--3365 (2010)
link

O. Pannekoucke.
Heterogeneous correlation modelling based on the wavelet diagonal assumption and
on the diffusion operator.
Mon. Wea. Rev. 137: 2995--3012 (2009). Special Issue on Mathematical Advances in Data Assimilation.
linkspecial issue

T. Lauvaux, O. Pannekoucke, C. Sarrat, F. Chevallier, P. Ciais,
J. Noilhan and P.J.O Rayner.
Structure of the transport uncertainty in mesoscale inversions of CO_2 sources
and sinks using ensemble model simulations.
Biogeosciences 6: 1089-1102 (2009).
link

O. Pannekoucke and S. Massart.
Estimation of the local diffusion tensor and normalization for heterogeneous correlation modelling
using a diffusion.
Q. J. R. Meteorol. Soc. 134: 1425--1438 (2008).
link

O. Pannekoucke, L. Berre and G. Desroziers.
Background error correlation length-scale estimates and their sampling statistics.
Q. J. R. Meteorol. Soc. 134: 497--508 (2008).
link

O. Pannekoucke, L. Berre and G. Desroziers,
Filtering properties of wavelets for the local background error correlations.
Q. J. R. Meteorol. Soc. 133: 363--379 (2007).
link

Proceedings

B. Pajot, S. Massart, D. Cariolle, A. Piacentini, O. Pannekoucke, W. Lahoz, C. Clerbaux, P. F. Coheur, and D. Hurtmans.
High resolution assimilation of IASI ozone data with a global CTM. In Concordiasi Workshop, Toulouse, France, Meteo-France/CNES

O. Pannekoucke, T. Lauvaux, C. Sarrat, P. Rayner, F. Chevallier et J. Noilhan.
Utilisation de previsions d'ensemble pour la modelisation des erreurs liees au transport applique à
l'inversion du CO2 a mesoechelle, "atelier de modelisation de l'atmosphere 2010",
Toulouse, du 26 au 28 janvier 2010.

How important is to use diagnosed background error covariances for the atmospheric ozone analysis? S. Massart, A. Piacentini, and O. Pannekoucke. 5th WMO SYMPOSIUM ON DATA ASSIMILATION
Melbourne, Australia, 5 - 9 October 2009.

G. Desroziers, L. Berre, O. Pannekoucke, S. Ecaterina Stefenescu, P. Brousseau, L. Auger, B. Chapnik and L. Raynaud.
Flow-dependent error covariances from variational assimilation ensembles on global and regional domains
HIRLAM Technical Report No. 68, July 2008.
(The SRNWP workshop on High resolution data assimilation with emphasis on the use of
moisture-related observations was arranged 21-23 March 2007 at the Museum of Work,
Norrkping, Sweden.)

L. Berre, O. Pannekoucke, G. Desroziers, S. E. Stefanescu, B. Chapnik, and L. Raynaud, 2007 : A variational assimilation ensemble and the spatial filtering of its error covariances : increase of sample size by local spatial averaging. Proceedings of the ECMWF Workshop on Flow-dependent aspects
of data assimilation, 11-13 June 2007, pages 151--168.
link

Reports/Books

O. Pannekoucke, E. Emili, and O. Thual.
Modelling of Local Length-Scale Dynamics and Isotropizing Deformations:
Formulation
in Natural Coordinate System Mathematical and Computational Approaches in Advancing
Modern Science and Engineering, Springer.
link

O. Pannekoucke.
Dynamique et modelisation de l'information dans les modeles meteorologique. Habilitation dissertation. Novembre 2012.

M. Farge, K. Schneider, O. Pannekoucke and R. Nguyen van Yen. 2011
Multiscale methods for fluid dynamics: fractals, self-similar random processes and wavelets.
Chapter in Handbook on environmental fluid dynamics",
Taylor and Francis (Publisher).

C. Baehr and O. Pannekoucke. Some Issues and results on the EnKF and particule filters for meteorological models.
chapter in Chaotic Systems: Theory and Applications;
C. H. Skiadas and I. Dimotikalis (Editors)
World Scientific (Publisher) Proceeding of the 2nd Chaotic Modeling and Simulation International
Conference 1 - 5 June 2009 Chania Crete Greece. (Chapter in Chaotic Systems: Theory and Applications )
pdf

O. Pannekoucke and C. Baehr.
Kalman Filters Family in Geoscience and Beyond. chapter in
link
Nova Science (Publisher).

O. Pannekoucke. Modelisation des structures locales de covariance des erreurs de prevision a l'aide des ondelettes. Ph.D dissertation. Mars 2008.
Ph.D dissertation
link

Talks

O. Pannekoucke, R. Menard, M. Bocquet, R. Fablet, A. Perrot, S. Ricci, O. Thual. Contribution of the parametric Kalman filter in practical and theoretical data assimilation, WCRP-WWRP Symposium on Data Assimilation and Reanalysis and 2021 ECMWF Annual Seminar on Observations. Virtual. 13-18 Sept. 2021

O. Pannekoucke, R. Fablet, S. Ricci, R. Menard, M. Bocquet, O. Thual. Design of the parametric Kalman filter dynamics : from the symbolic computation to the numerical integration., Climath WK2 "Big Data, Data Assimilation, Uncertainty quantification", Institute Henri Poincaré, Paris, 12-15 November, France, 2019.

O. Pannekoucke, S. Ricci, R. Menard, M. Bocquet, O. Thual. Parametric Kalman filter : toward an alternative to the EnKF? Adjoint Workshop on Sensitivity Analysis and Data Assimilation in Meteorology and Oceanography 1-6 July 2018, Aveiro, Portugal.

O. Pannekoucke, S. Ricci, R. Menard, M. Bocquet, O. Thual. Parametric Kalman filter : toward an alternative to the EnKF? Numerical model, predictability and data assimilation in weather, ocean and climate. A Symposium Honoring the Legacy of Anna Trevisan, Bologna, 17-20 October, 2017

O. Pannekoucke, S. Ricci, R. Menard, M. Bocquet, O. Thual. Parametric Kalman filter : toward an alternative to the EnKF? 12th International EnKF workshop June 12-14, 2017

O. Pannekoucke, E. Emili, O. Thual. Modeling of local length-scale dynamics and isotropizing deformations: formulation in natural coordinate system. AMMCS-CAIMS, 2015, June 7-12, Waterloo, Ohayo.

O. Pannekoucke, E. Emili, O. Thual, Modelling of local length-scale dynamics and isotropizing deformations: formulation in natural coordinate system. in World Weather Open Science Conference, Montreal, Canada, 2014.

R. Mechri, C. Ottle, O. Pannekoucke and Kallel, A. Genetic Particle Smoother Thermal Sharpener : Methodology and application to pseudo-observations, 1st international Conference on Advanced Technologies for Signal & Image Processing ATSIP’2014

Oger N., Pannekoucke O., Doeurenbecher, A. and Arbogast, Ph., Sensitivity of the KFS to the trajectory of reference. 9^th Workshop on Adjoint Model Applications in Dynamic Meteorology, 10-41 October 2011, Cefalu, Sicily, Italy.

Ricci S., Pannekoucke O., Ninove F. and Thual O., Emulation of a Kalman Filter algorithm on a diffusive flood wave propagation model. AGU Fall Meeting, August, 2011.

Pannekoucke O. Ateliers de Modélisation de l'Atmosphère, 26 - 28 Janvier 2010, Toulouse.

Pannekoucke O., Non separable diffusion and wavelet covariance model 9th EMS / 9th ECAM, 28 Sept. - 3 Oct. 2009, Toulouse.

Pannekoucke O., Berre L. and Desroziers G. 7th Adjoint Worshop on Ajoint Applications in Dynamic Meteorology, Innsbruck, Austria, 8-13 October 2006.

Invited talks

O. Pannekoucke, BIRS Workshop on Mathematical approaches for data assimilation of atmospheric constituents and inverse modeling website, Banff, Canada, October, 2020

O. Pannekoucke, Neural Network Generator: from the physics to the deep learningAI4OAC, Brest, 20-25 January, 2020

O. Pannekoucke, S. Ricci, R. Menard, M. Bocquet, O. Thual. Parametric Kalman filter : modelization of the model error? EGU 2019, presentation in session 'Inverse Problems, Data Assimilation and Uncertainty Quantification in Geoscience' (NP5.1/AS5.18/HS3.6/OS4.21) of the General Assembly of the European Geosciences Union, Vienna 7-12 April 2019. website

O. Pannekoucke, S. Ricci, R. Menard, M. Bocquet, O. Thual. Parametric Kalman filter : toward an alternative to the EnKF? Conférence - Colloque National d’Assimilation de Données Rennes, du 26 septembre au 28 septembre 2018.

O. Pannekoucke, Multi-scale issues in Data-assimilation for geophysical applications , Workshop on Multiscale Modeling and its Applications: From Weather and Climate Models to Models of Materials Defects, April, 25-29, 2016, Fields Institute.

O. Pannekoucke, « practical use of the length-scale » in Workshop « Theoretical aspects of ensemble data assimilation for the Earth system » Les Houches, 6—10 Avril 2015.

O. Pannekoucke « Optimisation pour la prévision numérique opérationnelle ». Présentation orale invitée. Journées Mathématiques de l'optimisation et de la décision (MODE) de la SMAI, Dijon, 28-30 Mars, 2012.

O. Pannekoucke. « The use of wavelets in data assimilation ». Conférence ondelette au CNMAC (congré national de mathématique appliquées et numérique, Société Brésilienne de Mathématique Appliquée et Numérique), 17-21 Septembre, Aguas de Lindoia, Bresil, 2012.

O. Pannekoucke. « Dynamics and modelling of information in geophysical model ». Working Group on PDE Control, CIRM, 5-9 Novembre, Marseille, France, 2012.

O. Pannekoucke, Journées de Statistiques Rennaise « Ensemble methods for variational data assimilation and forecasting», 20-21 Octobre 2011, Rennes, France.

O. Pannekoucke Thematic days on « Vortex @ Toulouse », IMFT Numerical Weather Prediction, IMFT, 27 Juin 2011, Toulouse, France.

O. Pannekoucke Participation au groupe de travail « WAVELET -- Multiresolution and wavelet techniques for plasma and fluid turbulence. » CEMRACS 2010, CIRM, Marseille, France.

O. Pannekoucke, seminar at Laboratoire de Météorologie Dynamique. Paris 25 Juin 2009.

O. Pannekoucke, Berre L. and Desroziers G. Présentation orale invité au workshop on Ensemble Methods in Meteorology and Oceanography. SAMA-IPSL, Paris 15-16 Mai 2008, France.

O. Pannekoucke, Berre L. and Desroziers G. Présentation orale invité au workshop on Mathematical Advancement in Geophysical Data Assimilation. Banff International Research Station for Mathematical Innovation and Discovery, Canada, 3-8 February 2008.

O. Pannekoucke, Berre L. and Desroziers G. Présentation orale invité au workshop on Flow-dependent aspects of data assimilation, ECMWF,Research Department, Shinfield Park, Reading, England, 11-13 June 2007.

Posters

O. Pannekoucke and Ronan Fablet. Automated NN generation: from the symbolic computation to the design of NN architectures for numerical predictions. ECMWF-ESA Virt. Workshop on ML for Earth System Observation and Prediction, 5-8 October 2020.

O. Pannekoucke, Presentation of the Parametric Kalman Filter (KAPA) Colloque de Bilan et de Prospective du programme LEFE, Clermont-Ferrand, 23-31 Mars 2018.

O. Pannekoucke, M. Bocquet, R. Menard. Parametric Covariance Propagation in the non-linear diffusive Burgers equation, 5th International Symposium on Data Assimilation, University of Reading, Reading, UK, 18-22 July 2016.

O. Pannekoucke, S. Ricci, S. Barthelemy, R. Menard, O. Thual. Parametric Kalman filter for Chemical Transport Models, 5th International Symposium on Data Assimilation, University of Reading, Reading, UK, 18-22 July 2016.

O. Pannekoucke, Berre L. and Desroziers G. Poster au Colloque National sur l'Assimilation de données, Toulouse, 9-10 mai 2006 .

Pannekoucke O., Berre L. and Desroziers G. Poster au fourth WMO International Symposium on Assimilation of Observations in Meteorology and Oceanography, Prague, Czech Republic, 18-22 April 2005.

Projects

Contribution in Scientific projects as principal investigator (PI), work package leader (WPL) or participant (P)

The project joins the CNRM (PI) and the CERFACS,
to explore the multivariate extension of the parametric Kalman filter (PKF).

The PKF is a method which reproduces the uncertainty evolution given by
the Kalman filter, but where the covariance matrices are approximated by
some covariance model tuned from a set of parameters.
In this project we will explore the ability of the PKF to apply for multivariate statistics.

The related contributions to the project are:

Research articles

M. Sabathier, O. Pannekoucke, V. Maget, and N. Dahmen, Boundary Conditions for the Parametric Kalman Filter forecast submited, 2022.. doi: 10.1002/essoar.10512724.1

A. Perrot, O. Pannekoucke and V. Guidard, Toward a multivariate formulation of the PKF ssimilation: application to a simplified chemical transport model. submitted egusphere-2022-928

(co-WPL) ANR JCJC PPOESY (Probabilitic prediction Of Extreme weather events with a
ai/physics SYnergy) (2021 – 2024)

In this project I contribute to explore the use of GANs approach applied for sampling AROME precipitation outputs, with the co-supervision of a postdoctoral fellowship. This contribution relies on the work done at CERFACS on WGANs (Besombes et al. 2021).

The project joins the CNRM (PI), the CERFACS, the CEREA
to explore the parametric Kalman filter (PKF).

The PKF is a method which reproduces the uncertainty evolution given by
the Kalman filter, but where the covariance matrices are approximated by
some covariance model tuned from a set of parameters.

The related contributions to the project are:

Research articles

O. Pannekoucke and P. Arbogast,
“SymPKF (v1.0): a symbolic and computational toolbox for the design of parametric Kalman filter dynamics,” Geosci. Model Dev., 14, 5957–5976, 2021
doi: 10.5194/gmd-14-5957-2021

R. Ménard, S. Skachko, and O. Pannekoucke,
“Numerical discretization causing error variance loss and the need for inflation,” Quarterly Journal of the Royal Meteorological Society, Aug. 2021,
doi: 10.1002/qj.4139

O. Pannekoucke,
An anisotropic formulation of the parametric Kalman filter assimilation,Tellus A: Dynamic Meteorology and Oceanography, vol. 73, no. 1, pp. 1–27, Jan. 2021, doi: 10.1080/16000870.2021.1926660.

O. Pannekoucke and R. Fablet,
“PDE-NetGen 1.0: from symbolic PDE representations of physical processes to trainable
neural network representations,” Geosci. Model Dev., 13, 3373–3382, 2020.
https://doi.org/10.5194/gmd-13-3373-2020

O. Pannekoucke Ricci, S.; Barthelemy, S.; Ménard, R. & Thual., O.
Parametric Kalman filter for chemical transport models - Corrigendum
Tellus A: Dynamic Meteorology and Oceanography, 70, 1-2 (2018)
link

O. Pannekoucke, S. Ricci, S. Barthelemy, R. Menard and O. Thual,
Parametric Kalman filter for Chemical Transport Models.
Tellus A, 68:31547, (2016).
https://doi.org/10.3402/tellusa.v68.31547

Talks

O. Pannekoucke, R. Menard, M. Bocquet, R. Fablet, A. Perrot, S. Ricci, O. Thual. Contribution of the parametric Kalman filter in practical and theoretical data assimilation, WCRP-WWRP Symposium on Data Assimilation and Reanalysis and 2021 ECMWF Annual Seminar on Observations. Virtual. 13-18 Sept. 2021

O. Pannekoucke, S. Ricci, R. Menard, M. Bocquet, O. Thual. Parametric Kalman filter : toward an alternative to the EnKF? Adjoint Workshop on Sensitivity Analysis and Data Assimilation in Meteorology and Oceanography 1-6 July 2018, Aveiro, Portugal.

O. Pannekoucke, S. Ricci, R. Menard, M. Bocquet, O. Thual. Parametric Kalman filter : toward an alternative to the EnKF? Numerical model, predictability and data assimilation in weather, ocean and climate. A Symposium Honoring the Legacy of Anna Trevisan, Bologna, 17-20 October, 2017

O. Pannekoucke, S. Ricci, R. Menard, M. Bocquet, O. Thual. Parametric Kalman filter : toward an alternative to the EnKF? 12th International EnKF workshop June 12-14, 2017

Invited Talks

O. Pannekoucke, BIRS Workshop on Mathematical approaches for data assimilation of atmospheric constituents and inverse modeling website, Banff, Canada, October, 2020

O. Pannekoucke, Neural Network Generator: from the physics to the deep learningAI4OAC, Brest, 20-25 January, 2020,

O. Pannekoucke, in session 'Inverse Problems, Data Assimilation and Uncertainty Quantification in Geosciences' (NP5.1/AS5.18/HS3.6/OS4.21) of the upcoming General Assembly of the European Geosciences Union, Vienna 7-12 April 2019.
website

O. Pannekoucke, S. Ricci, R. Menard, M. Bocquet, O. Thual. Parametric Kalman filter : toward an alternative to the EnKF? Conférence - Colloque National d’Assimilation de Données Rennes, du 26 septembre au 28 septembre 2018.

Summer School CEA-EDF-INRIA in Numerical Analysis "Data Assimilation in Numerical Simulations",
Paris, 26 juin-7 juillet 2006 :
O. Pannekoucke, L. Berre et G. Desroziers. Introduction to wavelets - Covariance matrix modelling using wavelets