Articles | Volume 31, issue 3
https://doi.org/10.5194/npg-31-303-2024
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/npg-31-303-2024
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Selecting and weighting dynamical models using data-driven approaches
Pierre Le Bras
CORRESPONDING AUTHOR
Laboratoire d’Océanographie Physique et Spatiale, IUEM, Univ Brest CNRS IRD Ifremer, Brest, France
IMT Atlantique, Lab-STICC, UMR CNRS 6285, 29238, Brest, France
Odyssey team project, INRIA IMT Atlantique CNRS, Brest, France
Florian Sévellec
Laboratoire d’Océanographie Physique et Spatiale, IUEM, Univ Brest CNRS IRD Ifremer, Brest, France
Odyssey team project, INRIA IMT Atlantique CNRS, Brest, France
Pierre Tandeo
IMT Atlantique, Lab-STICC, UMR CNRS 6285, 29238, Brest, France
Odyssey team project, INRIA IMT Atlantique CNRS, Brest, France
Juan Ruiz
Facultad de Ciencias Exactas y Naturales, Departamento de Ciencias de la Atmósfera y los Océanos, Universidad de Buenos Aires, Buenos Aires, Argentina
Centro de Investigaciones del Mar y la Atmósfera (CIMA), CONICET–Universidad de Buenos Aires, Buenos Aires, Argentina
Instituto Franco-Argentino para el Estudio del Clima y sus Impactos (IRL IFAECI), CNRS–IRD–CONICET–UBA, Buenos Aires, Argentina
Pierre Ailliot
Univ Brest, CNRS UMR 6205, Laboratoire de Mathematiques de Bretagne Atlantique, Brest, France
Related authors
Juan Ruiz, Pierre Ailliot, Thi Tuyet Trang Chau, Pierre Le Bras, Valérie Monbet, Florian Sévellec, and Pierre Tandeo
Geosci. Model Dev., 15, 7203–7220, https://doi.org/10.5194/gmd-15-7203-2022, https://doi.org/10.5194/gmd-15-7203-2022, 2022
Short summary
Short summary
We present a new approach to validate numerical simulations of the current climate. The method can take advantage of existing climate simulations produced by different centers combining an analog forecasting approach with data assimilation to quantify how well a particular model reproduces a sequence of observed values. The method can be applied with different observations types and is implemented locally in space and time significantly reducing the associated computational cost.
Juan Martin Guerrieri, Manuel Arturo Pulido, Takemasa Miyoshi, Arata Amemiya, and Juan José Ruiz
EGUsphere, https://doi.org/10.5194/egusphere-2025-2420, https://doi.org/10.5194/egusphere-2025-2420, 2025
Short summary
Short summary
This work extends the Mapping Particle Filter to account for local dependencies. Two localization methods are tested: a global particle flow with local kernels, and iterative local mappings based on correlation radius. Using a two-scale Lorenz-96 truth and a one-scale forecast model, experiments with full/partial and linear/nonlinear observations show Root Mean Square Error (RMSE) reductions using localized Gaussian mixture priors, achieving competitive performance against Gaussian filters.
Amélie Simon, Pierre Tandeo, Florian Sévellec, and Camille Lique
EGUsphere, https://doi.org/10.5194/egusphere-2025-704, https://doi.org/10.5194/egusphere-2025-704, 2025
Short summary
Short summary
This paper presents a new way to describe the Arctic sea-ice changes based on the shape of the observed seasonal cycles and using machine learning techniques. We show that the East Siberian and Laptev seas have lost their typical permanent sea-ice seasonal cycle while the Kara and Chukchi seas are experiencing a new typical seasonal cycle, corresponding to a partial winter-freezing.
Arthur Coquereau, Florian Sévellec, Thierry Huck, Joël J.-M. Hirschi, and Quentin Jamet
EGUsphere, https://doi.org/10.5194/egusphere-2025-17, https://doi.org/10.5194/egusphere-2025-17, 2025
Short summary
Short summary
Using statistical methods and a set of ensemble climate models, we decompose the sources of Atlantic Meridional Overturning Circulation (AMOC) variance. Three distinct phases of physical variability are identified: from 1850 to 1990, internal variability dominates; from 1990 to 2050, dynamical adjustment related to AMOC decline takes over; after 2050, differences between forcing scenarios become dominant. Beyond these physical factors, model variability remains the major source of uncertainty.
Paul Platzer, Pierre Ailliot, Bertrand Chapron, and Pierre Tandeo
Clim. Past, 20, 2267–2286, https://doi.org/10.5194/cp-20-2267-2024, https://doi.org/10.5194/cp-20-2267-2024, 2024
Short summary
Short summary
Old observations are necessary to understand the atmosphere. When direct observations are not available, one can use indirect observations, such as tide gauges, which measure the sea level in port cities. The sea level rises when local air pressure decreases and when wind pushes water towards the coast. Several centuries-long tide gauge records are available. We show that these can be complementary to direct pressure observations for studying storms and anticyclones in the 19th century.
Pierre Tandeo, Pierre Ailliot, and Florian Sévellec
Nonlin. Processes Geophys., 30, 129–137, https://doi.org/10.5194/npg-30-129-2023, https://doi.org/10.5194/npg-30-129-2023, 2023
Short summary
Short summary
The goal of this paper is to obtain probabilistic predictions of a partially observed dynamical system without knowing the model equations. It is illustrated using the three-dimensional Lorenz system, where only two components are observed. The proposed data-driven procedure is low-cost, is easy to implement, uses linear and Gaussian assumptions and requires only a small amount of data. It is based on an iterative linear Kalman smoother with a state augmentation.
Said Obakrim, Pierre Ailliot, Valérie Monbet, and Nicolas Raillard
Adv. Stat. Clim. Meteorol. Oceanogr., 9, 67–81, https://doi.org/10.5194/ascmo-9-67-2023, https://doi.org/10.5194/ascmo-9-67-2023, 2023
Short summary
Short summary
Ocean wave climate has a significant impact on human activities, and its understanding is of socioeconomic and environmental importance. In this study, we propose a statistical model that predicts wave heights in a location in the Bay of Biscay. The proposed method allows us to understand the spatiotemporal relationship between wind and waves and predicts well both wind seas and swells.
Robin Marcille, Maxime Thiébaut, Pierre Tandeo, and Jean-François Filipot
Wind Energ. Sci., 8, 771–786, https://doi.org/10.5194/wes-8-771-2023, https://doi.org/10.5194/wes-8-771-2023, 2023
Short summary
Short summary
A novel data-driven method is proposed to design an optimal sensor network for the reconstruction of offshore wind resources. Based on unsupervised learning of numerical weather prediction wind data, it provides a simple yet efficient method for the siting of sensors, outperforming state-of-the-art methods for this application. It is applied in the main French offshore wind energy development areas to provide guidelines for the deployment of floating lidars for wind resource assessment.
Juan Ruiz, Pierre Ailliot, Thi Tuyet Trang Chau, Pierre Le Bras, Valérie Monbet, Florian Sévellec, and Pierre Tandeo
Geosci. Model Dev., 15, 7203–7220, https://doi.org/10.5194/gmd-15-7203-2022, https://doi.org/10.5194/gmd-15-7203-2022, 2022
Short summary
Short summary
We present a new approach to validate numerical simulations of the current climate. The method can take advantage of existing climate simulations produced by different centers combining an analog forecasting approach with data assimilation to quantify how well a particular model reproduces a sequence of observed values. The method can be applied with different observations types and is implemented locally in space and time significantly reducing the associated computational cost.
Etienne Pauthenet, Loïc Bachelot, Kevin Balem, Guillaume Maze, Anne-Marie Tréguier, Fabien Roquet, Ronan Fablet, and Pierre Tandeo
Ocean Sci., 18, 1221–1244, https://doi.org/10.5194/os-18-1221-2022, https://doi.org/10.5194/os-18-1221-2022, 2022
Short summary
Short summary
Temperature and salinity profiles are essential for studying the ocean’s stratification, but there are not enough of these data. Satellites are able to measure daily maps of the surface ocean. We train a machine to learn the link between the satellite data and the profiles in the Gulf Stream region. We can then use this link to predict profiles at the high resolution of the satellite maps. Our prediction is fast to compute and allows us to get profiles at any locations only from surface data.
Marceau Michel, Said Obakrim, Nicolas Raillard, Pierre Ailliot, and Valérie Monbet
Adv. Stat. Clim. Meteorol. Oceanogr., 8, 83–95, https://doi.org/10.5194/ascmo-8-83-2022, https://doi.org/10.5194/ascmo-8-83-2022, 2022
Short summary
Short summary
In this study, we introduce a deep learning algorithm to establish the relationship between wind and waves in order to predict the latter. The performance of the proposed method has been evaluated both on the output of numerical wave models and on in situ data and compared to other statistical methods developed by our research team. The results obtained confirm the interest of deep learning methods for forecasting ocean data.
Juan Ruiz, Guo-Yuan Lien, Keiichi Kondo, Shigenori Otsuka, and Takemasa Miyoshi
Nonlin. Processes Geophys., 28, 615–626, https://doi.org/10.5194/npg-28-615-2021, https://doi.org/10.5194/npg-28-615-2021, 2021
Short summary
Short summary
Effective use of observations with numerical weather prediction models, also known as data assimilation, is a key part of weather forecasting systems. For precise prediction at the scales of thunderstorms, fast nonlinear processes pose a grand challenge because most data assimilation systems are based on linear processes and normal distribution errors. We investigate how, every 30 s, weather radar observations can help reduce the effect of nonlinear processes and nonnormal distributions.
Cited articles
Abramowitz, G., Herger, N., Gutmann, E., Hammerling, D., Knutti, R., Leduc, M., Lorenz, R., Pincus, R., and Schmidt, G. A.: ESD Reviews: Model dependence in multi-model climate ensembles: weighting, sub-selection and out-of-sample testing, Earth Syst. Dynam., 10, 91–105, https://doi.org/10.5194/esd-10-91-2019, 2019. a, b
Amos, M., Young, P. J., Hosking, J. S., Lamarque, J.-F., Abraham, N. L., Akiyoshi, H., Archibald, A. T., Bekki, S., Deushi, M., Jöckel, P., Kinnison, D., Kirner, O., Kunze, M., Marchand, M., Plummer, D. A., Saint-Martin, D., Sudo, K., Tilmes, S., and Yamashita, Y.: Projecting ozone hole recovery using an ensemble of chemistry–climate models weighted by model performance and independence, Atmos. Chem. Phys., 20, 9961–9977, https://doi.org/10.5194/acp-20-9961-2020, 2020. a
Brunner, L., Lorenz, R., Zumwald, M., and Knutti, R.: Quantifying uncertainty in European climate projections using combined performance-independence weighting, Environ. Res. Lett., 14, 124010, https://doi.org/10.1088/1748-9326/ab492f, 2019. a
Brunner, L., Pendergrass, A. G., Lehner, F., Merrifield, A. L., Lorenz, R., and Knutti, R.: Reduced global warming from CMIP6 projections when weighting models by performance and independence, Earth Syst. Dynam., 11, 995–1012, https://doi.org/10.5194/esd-11-995-2020, 2020. a
Carrassi, A., Bocquet, M., Bertino, L., and Evensen, G.: Data assimilation in the geosciences: An overview of methods, issues, and perspectives, Wires Clim. Change, 2018, e535, https://doi.org/10.1002/wcc.535, 2018. a
Cleveland, W. S. and Devlin, S. J.: Locally weighted regression: an approach to regression analysis by local fitting, J. Ame. Stat. Assoc., 83, 596–610, 1988. a
Dewar, W. K. and Huang, R. X.: Fluid flow in loops driven by freshwater and heat fluxes, J. Fluid Mech., 297, 153–191, 1995. a
Dewar, W. K. and Huang, R. X.: On the forced flow of salty water in a loop, Phys. Fluids, 8, 954–970, 1996. a
Evensen, G.: The ensemble Kalman filter: Theoretical formulation and practical implementation, Ocean Dynam., 53, 343–367, 2003. a
Eyring, V., Bony, S., Meehl, G. A., Senior, C. A., Stevens, B., Stouffer, R. J., and Taylor, K. E.: Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization, Geosci. Model Dev., 9, 1937–1958, https://doi.org/10.5194/gmd-9-1937-2016, 2016. a, b
Eyring, V., Cox, P. M., Flato, G. M., Gleckler, P. J., Abramowitz, G., Caldwell, P., Collins, W. D., Gier, B. K., Hall, A. D., Hoffman, F. M., Hurtt, G.C., Jahn, A., Chris D. Jones, C.D., Klein, S. A., Krasting, J. P., Kwiatkowski, L., Lorenz, R., Maloney, E., Meehl, G. A., Pendergrass, A. G., Pincus, R., Ruane, A. C., Russell, J. L., Sanderson, B. M., Santer, B. D., Sherwood, S. C., Simpson, I. R., Stouffer, R. J., and Williamson, M. S.: Taking climate model evaluation to the next level, Nat. Clim. Change, 9, 102–110, 2019. a
Garthwaite, P. and Mubwandarikwa, E.: Selection of weights for weighted model averaging: prior weights for weighted model averaging, Aust. N. Zeal. J. Stat., 52, 363–382, 2010. a
George, E. I.: Dilution priors: Compensating for model space redundancy, in: Borrowing Strength: Theory Powering Applications–A Festschrift for Lawrence D. Brown, Institute of Mathematical Statistics, vol. 6, 158–166, https://doi.org/10.1214/10-IMSCOLL611, 2010. a
Giorgi, F. and Mearns, L. O.: Calculation of average, uncertainty range, and reliability of regional climate changes from AOGCM simulations via the “reliability ensemble averaging” (REA) method, J. Climate, 15, 1141–1158, 2002. a
Herger, N., Abramowitz, G., Knutti, R., Angélil, O., Lehmann, K., and Sanderson, B. M.: Selecting a climate model subset to optimise key ensemble properties, Earth Syst. Dynam., 9, 135–151, https://doi.org/10.5194/esd-9-135-2018, 2018. a
Howard, L.: ABC's of convection, Geophysical Fluid Dynamics Summer School, WHOI Internal Tech. Rep, 71, 102–105, 1971. a
Huang, R. X. and Dewar, W. K.: Haline circulation: Bifurcation and chaos, J. Phys. Oceanogr., 26, 2093–2106, 1996. a
IPCC: Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, https://doi.org/10.1017/9781009157896, 2021. a
Jarque, C. M. and Bera, A. K.: Efficient tests for normality, homoscedasticity and serial independence of regression residuals, Econ. Lett., 6, 255–259, 1980. a
Knutti, R.: The end of model democracy?, Climatic Change, 102, 395–404, https://doi.org/10.1007/s10584-010-9800-2, 2010. a, b
Knutti, R., Furrer, R., Tebaldi, C., Cermak, J., and Meehl, G. A.: Challenges in combining projections from multiple climate models, J. Climate, 23, 2739–2758, 2010. a
Knutti, R., Baumberger, C., and Hadorn, G. H.: Uncertainty quantification using multiple models – Prospects and challenges, in: Computer Simulation Validation, Springer, 835–855, https://doi.org/10.1007/978-3-319-70766-2_34, 2019. a
Lorenz, E. N.: Atmospheric predictability as revealed by naturally occurring analogues, J. Atmos. Sci., 26, 636–646, 1969. a
Malkus, W. V.: Non-periodic convection at high and low Prandtl number, Mem. Societe Royale des Sciences de Liege, 125–128, 1972. a
Merrifield, A. L., Brunner, L., Lorenz, R., Medhaug, I., and Knutti, R.: An investigation of weighting schemes suitable for incorporating large ensembles into multi-model ensembles, Earth Syst. Dynam., 11, 807–834, https://doi.org/10.5194/esd-11-807-2020, 2020. a
Metref, S., Hannart, A., Ruiz, J., Bocquet, M., Carrassi, A., and Ghil, M.: Estimating model evidence using ensemble-based data assimilation with localization – The model selection problem, Q. J. Roy. Meteor. Soc., 145, 1571–1588, https://doi.org/10.1002/qj.3513, 2019. a, b, c
Min, S.-K., Simonis, D., and Hense, A.: Probabilistic climate change predictions applying Bayesian model averaging, Philos. T. Roy. Soc. A, 365, 2103–2116, 2007. a
Olson, R., Fan, Y., and Evans, J. P.: A simple method for Bayesian model averaging of regional climate model projections: Application to southeast Australian temperatures, Geophys. Res. Lett., 43, 7661–7669, 2016. a
Olson, R., An, S.-I., Fan, Y., and Evans, J. P.: Accounting for skill in trend, variability, and autocorrelation facilitates better multi-model projections: Application to the AMOC and temperature time series, Plos one, 14, e0214535, https://doi.org/10.1371/journal.pone.0214535, 2019. a
Perkins, S., Pitman, A., Holbrook, N. J., and McAneney, J.: Evaluation of the AR4 climate models’ simulated daily maximum temperature, minimum temperature, and precipitation over Australia using probability density functions, J. Climate, 20, 4356–4376, 2007. a
pilebras: pilebras/AnDA_weight_idealAMOC: AnDA_weight_ideal_AMOC_archived (AnDA_weight_AMOC), Zenodo [code and data set], https://doi.org/10.5281/zenodo.12575576, 2024. a
Pulido, M., Tandeo, P., Bocquet, M., Carrassi, A., and Lucini, M.: Stochastic parameterization identification using ensemble Kalman filtering combined with maximum likelihood methods, Tellus A, 70, 1–17, 2018. a
Raftery, A. E., Gneiting, T., Balabdaoui, F., and Polakowski, M.: Using Bayesian model averaging to calibrate forecast ensembles, Mon. Weather Rev., 133, 1155–1174, 2005. a
Ribes, A., Qasmi, S., and Gillett, N. P.: Making climate projections conditional on historical observations, Sci. Adv., 7, eabc0671, https://doi.org/10.1126/sciadv.abc0671, 2021. a
Ribes, A., Boé, J., Qasmi, S., Dubuisson, B., Douville, H., and Terray, L.: An updated assessment of past and future warming over France based on a regional observational constraint, Earth Syst. Dynam., 13, 1397–1415, https://doi.org/10.5194/esd-13-1397-2022, 2022. a
Ruiz, J., Ailliot, P., Chau, T. T. T., Le Bras, P., Monbet, V., Sévellec, F., and Tandeo, P.: Analog data assimilation for the selection of suitable general circulation models, Geosci. Model Dev., 15, 7203–7220, https://doi.org/10.5194/gmd-15-7203-2022, 2022. a, b, c
Sanderson, B. M., Wehner, M., and Knutti, R.: Skill and independence weighting for multi-model assessments, Geosci. Model Dev., 10, 2379–2395, https://doi.org/10.5194/gmd-10-2379-2017, 2017. a
Sévellec, F. and Fedorov, A. V.: Unstable AMOC during glacial intervals and millennial variability: The role of mean sea ice extent, Earth Planet. Sc. Lett., 429, 60–68, 2015. a
Sexton, D. M., Murphy, J. M., Collins, M., and Webb, M. J.: Multivariate probabilistic projections using imperfect climate models part I: outline of methodology, Clim. Dynam., 38, 2513–2542, 2012. a
Tandeo, P., Ailliot, P., Ruiz, J., Hannart, A., Chapron, B., Cuzol, A., Monbet, V., Easton, R., and Fablet, R.: Combining analog method and ensemble data assimilation: application to the Lorenz-63 chaotic system, in: Machine Learning and Data Mining Approaches to Climate Science: proceedings of the 4th International Workshop on Climate Informatics, Springer, 3–12, https://doi.org/10.1007/978-3-319-17220-0_1, 2015. a
Tandeo, P., Ailliot, P., Bocquet, M., Carrassi, A., Miyoshi, T., Pulido, M., and Zhen, Y.: A review of innovation-based methods to jointly estimate model and observation error covariance matrices in ensemble data assimilation, Mon. Weather Rev., 148, 3973–3994, 2020. a
Tebaldi, C. and Knutti, R.: The use of the multi-model ensemble in probabilistic climate projections, Philos. T. Roy. Soc. A, 365, 2053–2075, 2007. a
Van Leeuwen, P. J.: Particle filtering in geophysical systems, Mon. Weather Rev., 137, 4089–4114, 2009. a
Welander, P.: Note on the self-sustained oscillations of a simple thermal system, Tellus, 9, 419–420, 1957. a
Welander, P.: Steady and oscillatory motions of a differentially heated fluid loop, Woods Hole Oceanographic Institution, https://doi.org/10.1017/S0022112067000606, 1965. a
Welander, P.: On the oscillatory instability of a differentially heated fluid loop, J. Fluid Mech., 29, 17–30, 1967. a
Zhen, Y., Tandeo, P., Leroux, S., Metref, S., Penduff, T., and Le Sommer, J.: An adaptive optimal interpolation based on analog forecasting: application to SSH in the Gulf of Mexico, J. Atmos. Ocean. Tech., 37, 1697–1711, 2020. a
Short summary
The goal of this paper is to weight several dynamic models in order to improve the representativeness of a system. It is illustrated using a set of versions of an idealized model describing the Atlantic Meridional Overturning Circulation. The low-cost method is based on data-driven forecasts. It enables model performance to be evaluated on their dynamics. Taking into account both model performance and codependency, the derived weights outperform benchmarks in reconstructing a model distribution.
The goal of this paper is to weight several dynamic models in order to improve the...