{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,22]],"date-time":"2026-10-22T06:34:48Z","timestamp":1792650888518,"version":"4.5.0"},"reference-count":100,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2026,7,14]],"date-time":"2026-07-14T00:00:00Z","timestamp":1783987200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,7,14]],"date-time":"2026-07-14T00:00:00Z","timestamp":1783987200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/501100004815","name":"Isaac Newton Trust","doi-asserted-by":"publisher","award":["LEAG\/929"],"award-info":[{"award-number":["LEAG\/929"]}],"id":[{"id":"10.13039\/501100004815","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Nat Commun"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>\n                    Many systems resist analytical modeling, making data-driven inference of dynamics important. Yet data-driven methods can fail to converge or generalize, leaving open a central question:\n                    <jats:italic>When can system behavior be learned reliably from data, and when is such learning impossible?<\/jats:italic>\n                    We answer this question using adversarial dynamical systems to identify the boundary between accessible and inaccessible regimes. In Koopman operator learning, a leading framework for representing nonlinear dynamics through linear spectral objects, we design optimal data-driven spectral algorithms with convergence and certification guarantees under conditions arising broadly in physical systems. This yields a convergence theory for Koopman-operator approximations and resolves a longstanding open problem in Koopman spectral analysis. Conversely, by constructing adversarial systems, we prove matching impossibility results: without these conditions, no single-sequence limiting procedure can guarantee learning, regardless of data quality. These results sharply characterize when data-driven spectral learning can succeed and when it must fail. We validate the framework on oscillators, chaotic fluid flows and Arctic sea ice concentration forecasting. In the latter, we uncover hidden modes of Arctic sea ice decline, deliver long-range forecasts with geographic error bounds, and outperform state-of-the-art dynamical and deep learning models at substantially lower computational cost, enabling real-time deployment on standard CPUs.\n                  <\/jats:p>","DOI":"10.1038\/s41467-026-74220-8","type":"journal-article","created":{"date-parts":[[2026,7,14]],"date-time":"2026-07-14T09:02:51Z","timestamp":1784019771000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":1,"title":["Adversarial dynamical systems characterize when data-driven learning succeeds or fails"],"prefix":"10.1038","volume":"17","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-4964-9575","authenticated-orcid":false,"given":"Matthew J.","family":"Colbrook","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Igor","family":"Mezi\u0107","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alexei","family":"Stepanenko","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,7,14]]},"reference":[{"key":"74220_CR1","doi-asserted-by":"publisher","first-page":"436","DOI":"10.1038\/nature14539","volume":"521","author":"Y LeCun","year":"2015","unstructured":"LeCun, Y., Bengio, Y. & Hinton, G. Deep learning. Nature 521, 436\u2013444 (2015).","journal-title":"Nature"},{"key":"74220_CR2","doi-asserted-by":"publisher","first-page":"583","DOI":"10.1038\/s41586-021-03819-2","volume":"596","author":"J Jumper","year":"2021","unstructured":"Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583\u2013589 (2021).","journal-title":"Nature"},{"key":"74220_CR3","doi-asserted-by":"publisher","first-page":"1409","DOI":"10.1038\/s42003-024-07113-5","volume":"7","author":"L Abriata","year":"2024","unstructured":"Abriata, L. The Nobel Prize in Chemistry: Past, present, and future of AI in biology. Commun. Biol. 7, 1409 (2024).","journal-title":"Commun. Biol."},{"key":"74220_CR4","unstructured":"H., K., Zhang, X., Ren, S. & Sun, J. Deep residual learning for image recognition. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (IEEE, 2016)."},{"key":"74220_CR5","doi-asserted-by":"publisher","first-page":"1038","DOI":"10.1038\/s41587-019-0224-x","volume":"37","author":"A Zhavoronkov","year":"2019","unstructured":"Zhavoronkov, A. et al. Deep learning enables rapid identification of potent DDR1 kinase inhibitors. Nat. Biotechnol. 37, 1038\u20131040 (2019).","journal-title":"Nat. Biotechnol."},{"key":"74220_CR6","doi-asserted-by":"publisher","first-page":"715","DOI":"10.4310\/CMS.2003.v1.n4.a5","volume":"1","author":"CW Gear","year":"2003","unstructured":"Gear, C. W. et al. Equation-free, coarse-grained multiscale computation: enabling microscopic simulators to perform system-level analysis. Commun. Math. Sci. 1, 715\u2013762 (2003).","journal-title":"Commun. Math. Sci."},{"key":"74220_CR7","doi-asserted-by":"publisher","first-page":"309","DOI":"10.1007\/s11071-005-2824-x","volume":"41","author":"I Mezi\u0107","year":"2005","unstructured":"Mezi\u0107, I. Spectral properties of dynamical systems, model reduction and decompositions. Nonlinear Dyn. 41, 309\u2013325 (2005).","journal-title":"Nonlinear Dyn."},{"key":"74220_CR8","doi-asserted-by":"publisher","first-page":"81","DOI":"10.1126\/science.1165893","volume":"324","author":"M Schmidt","year":"2009","unstructured":"Schmidt, M. & Lipson, H. Distilling free-form natural laws from experimental data. Science 324, 81\u201385 (2009).","journal-title":"Science"},{"key":"74220_CR9","doi-asserted-by":"publisher","first-page":"032915","DOI":"10.1103\/PhysRevE.91.032915","volume":"91","author":"T Berry","year":"2015","unstructured":"Berry, T., Giannakis, D. & Harlim, J. Nonparametric forecasting of low-dimensional dynamical systems. Phys. Rev. E 91, 032915 (2015).","journal-title":"Phys. Rev. E"},{"key":"74220_CR10","doi-asserted-by":"publisher","first-page":"3932","DOI":"10.1073\/pnas.1517384113","volume":"113","author":"S Brunton","year":"2016","unstructured":"Brunton, S., Proctor, J. & Kutz, J. N. Discovering governing equations from data by sparse identification of nonlinear dynamical systems. Proc. Natl. Acad. Sci. 113, 3932\u20133937 (2016).","journal-title":"Proc. Natl. Acad. Sci."},{"key":"74220_CR11","doi-asserted-by":"crossref","unstructured":"Kutz, J. N., Brunton, S., Brunton, B. & Proctor, J.Dynamic Mode Decomposition: Data-Driven Modeling of Complex Systems (SIAM, 2016).","DOI":"10.1137\/1.9781611974508"},{"key":"74220_CR12","doi-asserted-by":"publisher","first-page":"338","DOI":"10.1016\/j.acha.2017.09.001","volume":"47","author":"D Giannakis","year":"2019","unstructured":"Giannakis, D. Data-driven spectral decomposition and forecasting of ergodic dynamical systems. Appl. Comput. Harmon. Anal. 47, 338\u2013396 (2019).","journal-title":"Appl. Comput. Harmon. Anal."},{"key":"74220_CR13","doi-asserted-by":"publisher","first-page":"477","DOI":"10.1146\/annurev-fluid-010719-060214","volume":"52","author":"S Brunton","year":"2020","unstructured":"Brunton, S., Noack, B. & Koumoutsakos, P. Machine learning for fluid mechanics. Annu. Rev. Fluid Mech. 52, 477\u2013508 (2020).","journal-title":"Annu. Rev. Fluid Mech."},{"key":"74220_CR14","doi-asserted-by":"publisher","first-page":"527","DOI":"10.1137\/21M144983X","volume":"22","author":"D Giannakis","year":"2023","unstructured":"Giannakis, D., Henriksen, A., Tropp, J. & Ward, R. Learning to forecast dynamical systems from streaming data. SIAM J. Appl. Dyn. Syst. 22, 527\u2013558 (2023).","journal-title":"SIAM J. Appl. Dyn. Syst."},{"key":"74220_CR15","doi-asserted-by":"publisher","first-page":"e2106140118","DOI":"10.1073\/pnas.2106140118","volume":"118","author":"T Bury","year":"2021","unstructured":"Bury, T. et al. Deep learning for early warning signals of tipping points. Proc. Natl. Acad. Sci. 118, e2106140118 (2021).","journal-title":"Proc. Natl. Acad. Sci."},{"key":"74220_CR16","doi-asserted-by":"publisher","first-page":"75","DOI":"10.1016\/j.acha.2021.02.004","volume":"54","author":"S Das","year":"2021","unstructured":"Das, S., Giannakis, D. & Slawinska, J. Reproducing kernel Hilbert space compactification of unitary evolution groups. Appl. Comput. Harmon. Anal. 54, 75\u2013136 (2021).","journal-title":"Appl. Comput. Harmon. Anal."},{"key":"74220_CR17","doi-asserted-by":"crossref","unstructured":"Gibney, E. China\u2019s cheap, open AI model DeepSeek thrills scientists. Nature 638, 13\u201314 (2025).","DOI":"10.1038\/d41586-025-00229-6"},{"key":"74220_CR18","doi-asserted-by":"publisher","first-page":"315","DOI":"10.1073\/pnas.17.5.315","volume":"17","author":"B Koopman","year":"1931","unstructured":"Koopman, B. Hamiltonian systems and transformation in Hilbert space. Proc. Natl. Acad. Sci. 17, 315\u2013318 (1931).","journal-title":"Proc. Natl. Acad. Sci."},{"key":"74220_CR19","doi-asserted-by":"publisher","first-page":"255","DOI":"10.1073\/pnas.18.3.255","volume":"18","author":"B Koopman","year":"1932","unstructured":"Koopman, B. & von Neumann, J. Dynamical systems of continuous spectra. Proc. Natl. Acad. Sci. 18, 255\u2013263 (1932).","journal-title":"Proc. Natl. Acad. Sci."},{"key":"74220_CR20","doi-asserted-by":"publisher","first-page":"047510","DOI":"10.1063\/1.4772195","volume":"22","author":"M Budi\u0161i\u0107","year":"2012","unstructured":"Budi\u0161i\u0107, M., Mohr, R. & Mezi\u0107, I. Applied Koopmanism. Chaos 22, 047510 (2012).","journal-title":"Chaos"},{"key":"74220_CR21","doi-asserted-by":"publisher","first-page":"229","DOI":"10.1137\/21M1401243","volume":"64","author":"S Brunton","year":"2022","unstructured":"Brunton, S., Budi\u0161i\u0107, M., Kaiser, E. & Kutz, J. N. Modern Koopman theory for dynamical systems. SIAM Rev. 64, 229\u2013340 (2022).","journal-title":"SIAM Rev."},{"key":"74220_CR22","first-page":"1","volume":"68","author":"I Mezi\u0107","year":"2021","unstructured":"Mezi\u0107, I. Koopman operator, geometry, and learning of dynamical systems. Not. Am. Math. Soc. 68, 1 (2021).","journal-title":"Not. Am. Math. Soc."},{"key":"74220_CR23","first-page":"127","volume":"25","author":"M Colbrook","year":"2024","unstructured":"Colbrook, M. The multiverse of dynamic mode decomposition algorithms. Handb. Numer. Anal. 25, 127\u2013230 (2024).","journal-title":"Handb. Numer. Anal."},{"key":"74220_CR24","doi-asserted-by":"publisher","first-page":"eadd6864","DOI":"10.1126\/scirobotics.add6864","volume":"8","author":"D Haggerty","year":"2023","unstructured":"Haggerty, D. et al. Control of soft robots with inertial dynamics. Sci. Robot. 8, eadd6864 (2023).","journal-title":"Sci. Robot."},{"key":"74220_CR25","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-021-26357-x","volume":"12","author":"G Froyland","year":"2021","unstructured":"Froyland, G., Giannakis, D., Lintner, B., Pike, M. & Slawinska, J. Spectral analysis of climate dynamics with operator-theoretic approaches. Nat. Commun. 12, 6570 (2021).","journal-title":"Nat. Commun."},{"key":"74220_CR26","unstructured":"Orvieto, A. et al. Resurrecting recurrent neural networks for long sequences. In International Conference on Machine Learning, 26670\u201326698 (PMLR, 2023)."},{"key":"74220_CR27","doi-asserted-by":"publisher","first-page":"139","DOI":"10.1093\/inthealth\/ihv009","volume":"7","author":"J Proctor","year":"2015","unstructured":"Proctor, J. & Eckhoff, P. Discovering dynamic patterns from infectious disease data using dynamic mode decomposition. Int. Health 7, 139\u2013145 (2015).","journal-title":"Int. Health"},{"key":"74220_CR28","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1016\/j.jneumeth.2015.10.010","volume":"258","author":"B Brunton","year":"2016","unstructured":"Brunton, B., Johnson, L., Ojemann, J. & Kutz, J. N. Extracting spatial-temporal coherent patterns in large-scale neural recordings using dynamic mode decomposition. J. Neurosci. Methods 258, 1\u201315 (2016).","journal-title":"J. Neurosci. Methods"},{"key":"74220_CR29","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-024-48033-6","volume":"15","author":"G Froyland","year":"2024","unstructured":"Froyland, G., Giannakis, D., Luna, E. & Slawinska, J. Revealing trends and persistent cycles of non-autonomous systems with autonomous operator-theoretic techniques. Nat. Commun. 15, 4268 (2024).","journal-title":"Nat. Commun."},{"key":"74220_CR30","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41467-018-07210-0","volume":"9","author":"B Lusch","year":"2018","unstructured":"Lusch, B., Kutz, J. N. & Brunton, S. Deep learning for universal linear embeddings of nonlinear dynamics. Nat. Commun. 9, 1\u201310 (2018).","journal-title":"Nat. Commun."},{"key":"74220_CR31","doi-asserted-by":"publisher","first-page":"687","DOI":"10.1007\/s00332-017-9423-0","volume":"28","author":"M Korda","year":"2018","unstructured":"Korda, M. & Mezi\u0107, I. On convergence of extended dynamic mode decomposition to the Koopman operator. J. Nonlinear Sci. 28, 687\u2013710 (2018).","journal-title":"J. Nonlinear Sci."},{"key":"74220_CR32","doi-asserted-by":"publisher","first-page":"5","DOI":"10.1017\/S0022112010001217","volume":"656","author":"P Schmid","year":"2010","unstructured":"Schmid, P. Dynamic mode decomposition of numerical and experimental data. J. Fluid Mech. 656, 5\u201328 (2010).","journal-title":"J. Fluid Mech."},{"key":"74220_CR33","doi-asserted-by":"publisher","first-page":"1307","DOI":"10.1007\/s00332-015-9258-5","volume":"25","author":"M Williams","year":"2015","unstructured":"Williams, M., Kevrekidis, I. & Rowley, C. A data-driven approximation of the Koopman operator: Extending dynamic mode decomposition. J. Nonlinear Sci. 25, 1307\u20131346 (2015).","journal-title":"J. Nonlinear Sci."},{"key":"74220_CR34","doi-asserted-by":"crossref","unstructured":"Babu\u0161ka, I. & Osborn, J. Eigenvalue problems. In Handbook of Numerical Analysis, vol. 2, 641\u2013787 (Elsevier, 1991).","DOI":"10.1016\/S1570-8659(05)80042-0"},{"key":"74220_CR35","doi-asserted-by":"crossref","unstructured":"Chatelin, F.Spectral Approximation of Linear Operators (Society for Industrial and Applied Mathematics, 2011).","DOI":"10.1137\/1.9781611970678"},{"key":"74220_CR36","doi-asserted-by":"publisher","first-page":"599","DOI":"10.1016\/j.acha.2018.08.002","volume":"48","author":"M Korda","year":"2020","unstructured":"Korda, M., Putinar, M. & Mezi\u0107, I. Data-driven spectral analysis of the Koopman operator. Appl. Comput. Harmon. Anal. 48, 599\u2013629 (2020).","journal-title":"Appl. Comput. Harmon. Anal."},{"key":"74220_CR37","doi-asserted-by":"publisher","first-page":"221","DOI":"10.1002\/cpa.22125","volume":"77","author":"M Colbrook","year":"2023","unstructured":"Colbrook, M. & Townsend, A. Rigorous data-driven computation of spectral properties of Koopman operators for dynamical systems. Commun. Pure Appl. Math. 77, 221\u2013283 (2023).","journal-title":"Commun. Pure Appl. Math."},{"key":"74220_CR38","doi-asserted-by":"publisher","first-page":"168","DOI":"10.1038\/s43247-022-00498-3","volume":"3","author":"M Rantanen","year":"2022","unstructured":"Rantanen, M. et al. The Arctic has warmed nearly four times faster than the globe since 1979. Commun. Earth Environ. 3, 168 (2022).","journal-title":"Commun. Earth Environ."},{"key":"74220_CR39","doi-asserted-by":"publisher","first-page":"20","DOI":"10.1038\/s41558-019-0662-y","volume":"10","author":"J Cohen","year":"2020","unstructured":"Cohen, J. et al. Divergent consensuses on Arctic amplification influence on midlatitude severe winter weather. Nat. Clim. Change 10, 20\u201329 (2020).","journal-title":"Nat. Clim. Change"},{"key":"74220_CR40","doi-asserted-by":"publisher","first-page":"1116","DOI":"10.1126\/science.abi9167","volume":"373","author":"J Cohen","year":"2021","unstructured":"Cohen, J. et al. Linking arctic variability and change with extreme winter weather in the United States. Science 373, 1116\u20131121 (2021).","journal-title":"Science"},{"key":"74220_CR41","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-018-05256-8","volume":"9","author":"D Coumou","year":"2018","unstructured":"Coumou, D. et al. The influence of Arctic amplification on mid-latitude summer circulation. Nat. Commun. 9, 2959 (2018).","journal-title":"Nat. Commun."},{"key":"74220_CR42","doi-asserted-by":"publisher","first-page":"3298","DOI":"10.1029\/2018GL081565","volume":"46","author":"N Wayand","year":"2019","unstructured":"Wayand, N., Bitz, C. & Blanchard-Wrigglesworth, E. A year-round subseasonal-to-seasonal sea ice prediction portal. Geophys. Res. Lett. 46, 3298\u20133307 (2019).","journal-title":"Geophys. Res. Lett."},{"key":"74220_CR43","doi-asserted-by":"publisher","first-page":"123801","DOI":"10.1103\/PhysRevFluids.9.123801","volume":"9","author":"J Lohmann","year":"2024","unstructured":"Lohmann, J. & Lucarini, V. Melancholia states of the Atlantic Meridional Overturning Circulation. Phys. Rev. Fluids 9, 123801 (2024).","journal-title":"Phys. Rev. Fluids"},{"key":"74220_CR44","doi-asserted-by":"publisher","first-page":"eadi4253","DOI":"10.1126\/sciadv.adi4253","volume":"10","author":"J Lohmann","year":"2024","unstructured":"Lohmann, J., Dijkstra, H., Jochum, M., Lucarini, V. & Ditlevsen, P. Multistability and intermediate tipping of the Atlantic Ocean circulation. Sci. Adv. 10, eadi4253 (2024).","journal-title":"Sci. Adv."},{"key":"74220_CR45","unstructured":"Mezi\u0107, I. Operator is the model. arXiv:2310.18516 (2023)."},{"key":"74220_CR46","unstructured":"Zekri, O. et al. Large language models as markov chains. arXiv preprint arXiv:2410.02724 (2024)."},{"key":"74220_CR47","unstructured":"Ma\u00f1\u00e9, R. Ergodic theory and differentiable dynamics, vol. 8 (2012)."},{"key":"74220_CR48","unstructured":"Mezi\u0107, I. On the geometrical and statistical properties of dynamical systems: Theory and applications. Ph.D. thesis (1994)."},{"key":"74220_CR49","doi-asserted-by":"publisher","first-page":"075021","DOI":"10.1088\/1361-6544\/ad4ade","volume":"37","author":"D Giannakis","year":"2024","unstructured":"Giannakis, D. & Valva, C. Consistent spectral approximation of Koopman operators using resolvent compactification. Nonlinearity 37, 075021 (2024).","journal-title":"Nonlinearity"},{"key":"74220_CR50","doi-asserted-by":"publisher","first-page":"19","DOI":"10.1016\/j.physd.2013.06.004","volume":"261","author":"A Mauroy","year":"2013","unstructured":"Mauroy, A., Mezi\u0107, I. & Moehlis, J. Isostables, isochrons, and Koopman spectrum for the action\u2013angle representation of stable fixed point dynamics. Phys. D: Nonlinear Phenom. 261, 19\u201330 (2013).","journal-title":"Phys. D: Nonlinear Phenom."},{"key":"74220_CR51","doi-asserted-by":"publisher","first-page":"1150","DOI":"10.1137\/24M1662370","volume":"24","author":"M Colbrook","year":"2025","unstructured":"Colbrook, M., Drysdale, C. & Horning, A. Rigged dynamic mode decomposition. SIAM J. Appl. Dyn. Syst. 24, 1150\u20131190 (2025).","journal-title":"SIAM J. Appl. Dyn. Syst."},{"key":"74220_CR52","doi-asserted-by":"publisher","first-page":"41","DOI":"10.1063\/1.2995142","volume":"31","author":"H Swinney","year":"1978","unstructured":"Swinney, H. & Gollub, J. The transition to turbulence. Phys. today 31, 41\u201349 (1978).","journal-title":"Phys. today"},{"key":"74220_CR53","doi-asserted-by":"publisher","first-page":"13","DOI":"10.1007\/BF00392202","volume":"124","author":"H Broer","year":"1993","unstructured":"Broer, H. & Takens, F. Mixed spectra and rotational symmetry. Arch. Ration. Mech. Anal. 124, 13\u201342 (1993).","journal-title":"Arch. Ration. Mech. Anal."},{"key":"74220_CR54","doi-asserted-by":"publisher","first-page":"5107","DOI":"10.1016\/j.jfa.2017.01.021","volume":"272","author":"J Fillman","year":"2017","unstructured":"Fillman, J. & Ong, D. Purely singular continuous spectrum for limit-periodic CMV operators with applications to quantum walks. J. Funct. Anal. 272, 5107\u20135143 (2017).","journal-title":"J. Funct. Anal."},{"key":"74220_CR55","doi-asserted-by":"publisher","first-page":"13490","DOI":"10.1073\/pnas.1508622112","volume":"112","author":"M Stuecker","year":"2015","unstructured":"Stuecker, M., Jin, F. & Timmermann, A. El Ni\u00f1o- southern oscillation frequency cascade. Proc. Natl. Acad. Sci. 112, 13490\u201313495 (2015).","journal-title":"Proc. Natl. Acad. Sci."},{"key":"74220_CR56","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-020-73211-z","volume":"10","author":"J Hogg","year":"2020","unstructured":"Hogg, J., Fonoberova, M. & Mezi\u0107, I. Exponentially decaying modes and long-term prediction of sea ice concentration using Koopman mode decomposition. Sci. Rep. 10, 16313 (2020).","journal-title":"Sci. Rep."},{"key":"74220_CR57","doi-asserted-by":"publisher","first-page":"1005","DOI":"10.1007\/s10584-011-0101-1","volume":"110","author":"J Stroeve","year":"2012","unstructured":"Stroeve, J. et al. The Arctic\u2019s rapidly shrinking sea ice cover: a research synthesis. Clim. change 110, 1005\u20131027 (2012).","journal-title":"Clim. change"},{"key":"74220_CR58","doi-asserted-by":"publisher","first-page":"869","DOI":"10.1038\/ngeo2277","volume":"7","author":"M Mori","year":"2014","unstructured":"Mori, M., Watanabe, M., Shiogama, H., Inoue, J. & Kimoto, M. Robust Arctic sea-ice influence on the frequent Eurasian cold winters in past decades. Nat. Geosci. 7, 869\u2013873 (2014).","journal-title":"Nat. Geosci."},{"key":"74220_CR59","doi-asserted-by":"publisher","first-page":"231","DOI":"10.1175\/2010JCLI3775.1","volume":"24","author":"E Blanchard-Wrigglesworth","year":"2011","unstructured":"Blanchard-Wrigglesworth, E., Armour, K., Bitz, C. & DeWeaver, E. Persistence and inherent predictability of Arctic sea ice in a GCM ensemble and observations. J. Clim. 24, 231\u2013250 (2011).","journal-title":"J. Clim."},{"key":"74220_CR60","doi-asserted-by":"publisher","first-page":"4657","DOI":"10.1175\/JCLI-D-16-0549.1","volume":"30","author":"M Bushuk","year":"2017","unstructured":"Bushuk, M. & Giannakis, D. The seasonality and interannual variability of Arctic sea ice reemergence. J. Clim. 30, 4657\u20134676 (2017).","journal-title":"J. Clim."},{"key":"74220_CR61","doi-asserted-by":"publisher","first-page":"6265","DOI":"10.1175\/JCLI-D-13-00256.1","volume":"27","author":"M Bushuk","year":"2014","unstructured":"Bushuk, M., Giannakis, D. & Majda, A. Reemergence mechanisms for North Pacific sea ice revealed through nonlinear Laplacian spectral analysis. J. Clim. 27, 6265\u20136287 (2014).","journal-title":"J. Clim."},{"key":"74220_CR62","doi-asserted-by":"publisher","first-page":"634","DOI":"10.1038\/s41558-018-0205-y","volume":"8","author":"S Lind","year":"2018","unstructured":"Lind, S., Ingvaldsen, R. & Furevik, T. Arctic warming hotspot in the northern Barents Sea linked to declining sea-ice import. Nat. Clim. Change 8, 634\u2013639 (2018).","journal-title":"Nat. Clim. Change"},{"key":"74220_CR63","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-021-25257-4","volume":"12","author":"T Andersson","year":"2021","unstructured":"Andersson, T. et al. Seasonal Arctic sea ice forecasting with probabilistic deep learning. Nat. Commun. 12, 5124 (2021).","journal-title":"Nat. Commun."},{"key":"74220_CR64","doi-asserted-by":"publisher","first-page":"1087","DOI":"10.5194\/gmd-12-1087-2019","volume":"12","author":"S Johnson","year":"2019","unstructured":"Johnson, S. et al. SEAS5: the new ECMWF seasonal forecast system. Geosci. Model Dev. 12, 1087\u20131117 (2019).","journal-title":"Geosci. Model Dev."},{"key":"74220_CR65","doi-asserted-by":"publisher","first-page":"430","DOI":"10.1038\/s41561-019-0363-1","volume":"12","author":"D Olonscheck","year":"2019","unstructured":"Olonscheck, D., Mauritsen, T. & Notz, D. Arctic sea-ice variability is primarily driven by atmospheric temperature fluctuations. Nat. Geosci. 12, 430\u2013434 (2019).","journal-title":"Nat. Geosci."},{"key":"74220_CR66","doi-asserted-by":"crossref","unstructured":"Arnold, V. I.Mathematical Methods of Classical Mechanics (Springer New York, 1989).","DOI":"10.1007\/978-1-4757-2063-1"},{"key":"74220_CR67","unstructured":"Hill, T.An introduction to statistical thermodynamics (Courier Corporation, 1986)."},{"key":"74220_CR68","doi-asserted-by":"publisher","first-page":"653","DOI":"10.1070\/RM1996v051n04ABEH002964","volume":"51","author":"A Kachurovskii","year":"1996","unstructured":"Kachurovskii, A. The rate of convergence in ergodic theorems. Russian Math. Surv. 51, 653\u2013703 (1996).","journal-title":"Russian Math. Surv."},{"key":"74220_CR69","unstructured":"Krengel, U. Ergodic theorems (Walter de Gruyter, 2011)."},{"key":"74220_CR70","doi-asserted-by":"crossref","unstructured":"Trefethen, L. N. & Embree, M. Spectra and Pseudospectra: The Behavior of Nonnormal Matrices and Operators (PUP, 2005).","DOI":"10.1515\/9780691213101"},{"key":"74220_CR71","doi-asserted-by":"publisher","first-page":"2235","DOI":"10.1063\/1.1480266","volume":"14","author":"I Mezi\u0107","year":"2002","unstructured":"Mezi\u0107, I. & Sotiropoulos, F. Ergodic theory and experimental visualization of invariant sets in chaotically advected flows. Phys. Fluids 14, 2235\u20132243 (2002).","journal-title":"Phys. Fluids"},{"key":"74220_CR72","doi-asserted-by":"publisher","first-page":"81","DOI":"10.1090\/S0894-0347-2010-00676-5","volume":"24","author":"A Hansen","year":"2011","unstructured":"Hansen, A. On the solvability complexity index, the n-pseudospectrum and approximations of spectra of operators. J. Am. Math. Soc. 24, 81\u2013124 (2011).","journal-title":"J. Am. Math. Soc."},{"key":"74220_CR73","unstructured":"Ben-Artzi, J., Colbrook, M., Hansen, A., Nevanlinna, O. & Seidel, M. Computing spectra - On the solvability complexity index hierarchy and towers of algorithms. arXiv 1508.03280. (2020)."},{"key":"74220_CR74","doi-asserted-by":"publisher","first-page":"671","DOI":"10.3934\/dcds.2007.17.671","volume":"17","author":"G Froyland","year":"2007","unstructured":"Froyland, G. On Ulam approximation of the isolated spectrum and eigenfunctions of hyperbolic maps. Discret. Continuous Dyn. Syst. 17, 671\u2013689 (2007).","journal-title":"Discret. Continuous Dyn. Syst."},{"key":"74220_CR75","doi-asserted-by":"publisher","first-page":"1905","DOI":"10.1088\/0951-7715\/15\/6\/309","volume":"15","author":"M Blank","year":"2002","unstructured":"Blank, M., Keller, G. & Liverani, C. Ruelle\u2013Perron\u2013Frobenius spectrum for Anosov maps. Nonlinearity 15, 1905\u20131973 (2002).","journal-title":"Nonlinearity"},{"key":"74220_CR76","doi-asserted-by":"publisher","first-page":"985","DOI":"10.1007\/s00332-017-9437-7","volume":"28","author":"S Klus","year":"2018","unstructured":"Klus, S. et al. Data-driven model reduction and transfer operator approximation. J. Nonlinear Sci. 28, 985\u20131010 (2018).","journal-title":"J. Nonlinear Sci."},{"key":"74220_CR77","doi-asserted-by":"publisher","first-page":"184","DOI":"10.1016\/j.automatica.2019.05.016","volume":"106","author":"S Peitz","year":"2019","unstructured":"Peitz, S. & Klus, S. Koopman operator-based model reduction for switched-system control of PDEs. Automatica 106, 184\u2013191 (2019).","journal-title":"Automatica"},{"key":"74220_CR78","doi-asserted-by":"crossref","unstructured":"Katznelson, Y.An Introduction to Harmonic Analysis (CUP, 2004).","DOI":"10.1017\/CBO9781139165372"},{"key":"74220_CR79","doi-asserted-by":"publisher","first-page":"357","DOI":"10.1146\/annurev-fluid-011212-140652","volume":"45","author":"I Mezi\u0107","year":"2013","unstructured":"Mezi\u0107, I. Analysis of fluid flows via spectral properties of the Koopman operator. Annu. Rev. Fluid Mech. 45, 357\u2013378 (2013).","journal-title":"Annu. Rev. Fluid Mech."},{"key":"74220_CR80","doi-asserted-by":"publisher","first-page":"461","DOI":"10.1016\/S0370-1573(02)00331-9","volume":"371","author":"G Zaslavsky","year":"2002","unstructured":"Zaslavsky, G. Chaos, fractional kinetics, and anomalous transport. Phys. Rep. 371, 461\u2013580 (2002).","journal-title":"Phys. Rep."},{"key":"74220_CR81","volume-title":"Nonlinear Time Series Analysis","author":"H Kantz","year":"2006","unstructured":"Kantz, H., Schreiber, T. Nonlinear Time Series Analysis. CUP: Cambridge, 2006."},{"key":"74220_CR82","doi-asserted-by":"publisher","first-page":"101","DOI":"10.1016\/j.physd.2004.06.015","volume":"197","author":"I Mezi\u0107","year":"2004","unstructured":"Mezi\u0107, I. & Banaszuk, A. Comparison of systems with complex behavior. Phys. D: Nonlinear Phenom. 197, 101\u2013133 (2004).","journal-title":"Phys. D: Nonlinear Phenom."},{"key":"74220_CR83","doi-asserted-by":"publisher","first-page":"744","DOI":"10.1038\/s42254-023-00650-8","volume":"5","author":"V Lucarini","year":"2023","unstructured":"Lucarini, V. & Chekroun, M. Theoretical tools for understanding the climate crisis from Hasselmann\u2019s programme and beyond. Nat. Rev. Phys. 5, 744\u2013765 (2023).","journal-title":"Nat. Rev. Phys."},{"key":"74220_CR84","unstructured":"Chen, R., Rubanova, Y., Bettencourt, J. & Duvenaud, D. Neural ordinary differential equations. Advances in neural information processing systems31 (2018)."},{"key":"74220_CR85","unstructured":"L., Z. et al. Fourier neural operator for parametric partial differential equations. In 9th International Conference on Learning Representations, ICLR 2021, Virtual Event, Austria, May 3-7, 2021 (OpenReview.net, 2021)."},{"key":"74220_CR86","doi-asserted-by":"publisher","first-page":"20170844","DOI":"10.1098\/rspa.2017.0844","volume":"474","author":"P Vlachas","year":"2018","unstructured":"Vlachas, P., Byeon, W., Wan, Z., Sapsis, T. & Koumoutsakos, P. Data-driven forecasting of high-dimensional chaotic systems with long short-term memory networks. Proc. R. Soc. A 474, 20170844 (2018).","journal-title":"Proc. R. Soc. A"},{"key":"74220_CR87","unstructured":"Long, Z., Lu, Y., Ma, X. & Dong, B. Pde-net: Learning PDEs from data. In International conference on machine learning, 3208\u20133216 (PMLR, 2018)."},{"key":"74220_CR88","doi-asserted-by":"publisher","first-page":"e2303904120","DOI":"10.1073\/pnas.2303904120","volume":"120","author":"N Boull\u00e9","year":"2023","unstructured":"Boull\u00e9, N., Halikias, D. & Townsend, A. Elliptic PDE learning is provably data-efficient. Proc. Natl. Acad. Sci. 120, e2303904120 (2023).","journal-title":"Proc. Natl. Acad. Sci."},{"key":"74220_CR89","doi-asserted-by":"publisher","first-page":"7228","DOI":"10.1109\/TPWRS.2018.2846744","volume":"33","author":"M Netto","year":"2018","unstructured":"Netto, M. & Mili, L. A robust data-driven Koopman Kalman filter for power systems dynamic state estimation. IEEE Trans. Power Syst. 33, 7228\u20137237 (2018).","journal-title":"IEEE Trans. Power Syst."},{"key":"74220_CR90","doi-asserted-by":"crossref","unstructured":"Arbabi, H., Korda, M. & Mezi\u0107, I. A data-driven Koopman model predictive control framework for nonlinear partial differential equations. In 2018 IEEE CDC, 6409\u20136414 (2018).","DOI":"10.1109\/CDC.2018.8619720"},{"key":"74220_CR91","doi-asserted-by":"publisher","first-page":"e16743","DOI":"10.1002\/aic.16743","volume":"65","author":"A Narasingam","year":"2019","unstructured":"Narasingam, A. & Kwon, J. Koopman Lyapunov-based model predictive control of nonlinear chemical process systems. AIChE J. 65, e16743 (2019).","journal-title":"AIChE J."},{"key":"74220_CR92","doi-asserted-by":"crossref","unstructured":"Hasnain, A., Boddupalli, N., Balakrishnan, S. & Yeung, E. Steady state programming of controlled nonlinear systems via deep dynamic mode decomposition. In IEEE ACC, 4245\u20134251 (2020).","DOI":"10.23919\/ACC45564.2020.9147218"},{"key":"74220_CR93","doi-asserted-by":"publisher","first-page":"101035","DOI":"10.1016\/j.arcontrol.2025.101035","volume":"61","author":"R Str\u00e4sser","year":"2026","unstructured":"Str\u00e4sser, R. et al. An overview of Koopman-based control: From error bounds to closed-loop guarantees. Annu. Rev. Control 61, 101035 (2026).","journal-title":"Annu. Rev. Control"},{"key":"74220_CR94","doi-asserted-by":"publisher","first-page":"124402","DOI":"10.1103\/PhysRevFluids.2.124402","volume":"2","author":"H Arbabi","year":"2017","unstructured":"Arbabi, H. & Mezi\u0107, I. Study of dynamics in post-transient flows using Koopman mode decomposition. Phys. Rev. Fluids 2, 124402 (2017).","journal-title":"Phys. Rev. Fluids"},{"key":"74220_CR95","unstructured":"Temam, R.Infinite-dimensional dynamical systems in mechanics and physics, vol. 68 of Applied Mathematical Sciences (Springer-Verlag, New York, 2012)."},{"key":"74220_CR96","doi-asserted-by":"crossref","unstructured":"Trefethen, L. N.Spectral methods in MATLAB (SIAM, 2000).","DOI":"10.1137\/1.9780898719598"},{"key":"74220_CR97","doi-asserted-by":"publisher","first-page":"433","DOI":"10.1007\/s00220-021-04072-4","volume":"384","author":"M Colbrook","year":"2021","unstructured":"Colbrook, M. Computing spectral measures and spectral types. Commun. Math. Phys. 384, 433\u2013501 (2021).","journal-title":"Commun. Math. Phys."},{"key":"74220_CR98","doi-asserted-by":"publisher","first-page":"1585","DOI":"10.1137\/22M1521407","volume":"61","author":"M Colbrook","year":"2023","unstructured":"Colbrook, M. The mpEDMD algorithm for data-driven computations of measure-preserving dynamical systems. SIAM J. Numer. Anal. 61, 1585\u20131608 (2023).","journal-title":"SIAM J. Numer. Anal."},{"key":"74220_CR99","doi-asserted-by":"publisher","first-page":"2096","DOI":"10.1137\/17M1125236","volume":"16","author":"H Arbabi","year":"2017","unstructured":"Arbabi, H. & Mezi\u0107, I. Ergodic theory, dynamic mode decomposition, and computation of spectral properties of the Koopman operator. SIAM J. Appl. Dyn. Syst. 16, 2096\u20132126 (2017).","journal-title":"SIAM J. Appl. Dyn. Syst."},{"key":"74220_CR100","doi-asserted-by":"publisher","first-page":"132416","DOI":"10.1016\/j.physd.2020.132416","volume":"406","author":"S Klus","year":"2020","unstructured":"Klus, S. et al. Data-driven approximation of the Koopman generator: Model reduction, system identification, and control. Phys. D: Nonlinear Phenom. 406, 132416 (2020).","journal-title":"Phys. D: Nonlinear Phenom."}],"container-title":["Nature Communications"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s41467-026-74220-8.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41467-026-74220-8","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41467-026-74220-8.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,14]],"date-time":"2026-07-14T09:02:55Z","timestamp":1784019775000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s41467-026-74220-8"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,14]]},"references-count":100,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,12]]}},"alternative-id":["74220"],"URL":"https:\/\/doi.org\/10.1038\/s41467-026-74220-8","relation":{},"ISSN":["2041-1723"],"issn-type":[{"value":"2041-1723","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,14]]},"assertion":[{"value":"20 February 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"28 May 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"14 July 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"The authors declare no competing interests.","order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}],"article-number":"5397"}}