Abstract
Data-driven deep learning (DL) models often underestimate the intensity of extreme weather and climate events due to the scarcity of extreme samples in training datasets and the smoothing effects of gradient-based optimization. While ensemble prediction methods based on initial condition (IC) perturbations in traditional numerical models have improved extreme event predictions, they often fail in DL frameworks. This is primarily due to limited error growth characteristics and the implicit regularization in DL models, which dampens the amplification of IC perturbations. To overcome this limitation, we introduce a novel IC perturbation scheme based on orthogonal conditional nonlinear optimal perturbation (O-CNOP), integrated into a DL-based ensemble prediction system. The O-CNOP-derived perturbations are obtained through an iterative selection and optimization process, beginning with candidate samples from model simulations under uniform energy constraints. Perturbations are then selected to maximize forecast error growth, guided by ensemble averaging and convergence criteria. We evaluate this method based on four major El Niño events (1982/83, 1997/98, 2015/16, and 2023/24). Results show significant improvements in DL model predictions when initialized in spring, with over a 30% reduction in prediction error for Niño3.4 sea surface temperature anomalies. This AI-enabled O-CNOP framework offers a robust and generalizable approach to ensemble predicting, potentially improving the prediction skill of DL-based weather and climate models for extreme events.
Similar content being viewed by others
Data availability
All data used in this study are publicly available online. CMIP6 products are available at https://aims2.llnl.gov/search. The GODAS reanalyses are available at https://iridl.ldeo.columbia.edu/SOURCES/.NOAA/.NCEP/.EMC/.CMB/.GODAS/.monthly/. The SODA reanalyses are available at https://soda.umd.edu/.
Code availability
The code, processed data, and model outputs can be found in https://zenodo.org/records/16986660 (https://doi.org/10.5281/zenodo.16986660).
References
Cai, W. et al. Pantropical climate interactions. Science 363, eaav4236 (2019).
Liu, Y., Cai, W., Lin, X., Li, Z. & Zhang, Y. Nonlinear El Niño impacts on the global economy under climate change. Nat. Commun. 14, 5887 (2023).
Zhang, R.-H., Rothstein, L. M. & Busalacchi, A. J. Origin of upper-ocean warming and El Niño change on decadal scales in the tropical Pacific Ocean. Nature 391, 879–883 (1998).
Timmermann, A. et al. El Niño-Southern Oscillation complexity. Nature 559, 535–545 (2018).
Thirumalai, K. et al. Future increase in extreme El Niño supported by past glacial changes. Nature 634, 374–380 (2024).
Rivera Tello, G. A., Takahashi, K. & Karamperidou, C. Explained predictions of strong eastern Pacific El Niño events using deep learning. Sci. Rep. 13, 21150 (2023).
Cane, M. A. & Zebiak, S. E. A theory for El Niño and the southern oscillation. Science 228, 1085–1087 (1985).
Cane, M. A., Zebiak, S. E. & Dolan, S. C. Experimental forecasts of El Niño. Nature 321, 827–832 (1986).
Chen, D., Zebiak, S. E., Busalacchi, A. J. & Cane, M. A. An Improved Procedure for EI Niño Forecasting: Implications for Predictability. Science 269, 1699–1702 (1995).
Zhang, R.-H., Zebiak, S. E., Kleeman, R. & Keenlyside, N. A new intermediate coupled model for El Niño simulation and prediction. Geophys. Res. Lett. 30, GL018010 (2003).
Barnston, A. G., Tippett, M. K., L’Heureux, M. L., Li, S. & DeWitt, D. G. Skill of Real-Time Seasonal ENSO Model Predictions during 2002-11: Is Our Capability Increasing?. Bull. Am. Meteorol. Soc. 93, 631–651 (2012).
Zhu, J., Wang, W., Kumar, A., Liu, Y. & DeWitt, D. Assessment of a New Global Ocean Reanalysis in ENSO Predictions With NOAA UFS. Geophys. Res. Lett. 51, e2023GL106640 (2024).
Ehsan, M. A., L’Heureux, M. L., Tippett, M. K., Robertson, A. W. & Turmelle, J. Real-time ENSO forecast skill evaluated over the last two decades, with focus on the onset of ENSO events. npj Clim. Atmos. Sci. 7, 301 (2024).
Ji, C., Mu, M., Fang, X. & Tao, L. Improving the Forecasting of El Niño Amplitude Based on an Ensemble Forecast Strategy for Westerly Wind Bursts. J. Clim. 36, 8675–8694 (2023).
Ji, C. et al. Toward skillful forecasting of super El Niño events using a diffusion-based westerly wind burst parameterization. npj Clim. Atmos. Sci. 8, 273 (2025).
Timmermann, A., Jin, F.-F. & Abshagen, J. A Nonlinear Theory for El Niño Bursting. J. Atmos. Sci. 60, 152–165 (2003).
Geng, L. & Jin, F.-F. Insights into ENSO Diversity from an Intermediate Coupled Model. Part II: Role of Nonlinear Dynamics and Stochastic Forcing. J. Clim. 36, 7527–7547 (2023).
Fang, X., Dijkstra, H., Wieners, C. & Guardamagna, F. A nonlinear full-field conceptual model for ENSO diversity. J. Clim. 37, 3759–3774 (2024).
Tippett, M. K., L’Heureux, M. L., Becker, E. J. & Kumar, A. Excessive Momentum and False Alarms in Late-Spring ENSO Forecasts. Geophys. Res. Lett. 47, e2020GL087008 (2020).
Jin, Y. S., Liu, Z. Y. & Duan, W. S. The Different Relationships between the ENSO Spring Persistence Barrier and Predictability Barrier. J. Clim. 35, 6207–6218 (2022).
Zhang, R.-H., Gao, C. & Feng, L. Recent ENSO evolution and its real-time prediction challenges. Natl. Sci. Rev. 9, nwac052 (2022).
Ham, Y. G., Kim, J. H. & Luo, J. J. Deep learning for multi-year ENSO forecasts. Nature 573, 568–572 (2019).
Zhang, R.-H., Zhou, L., Gao, C. & Tao, L. A transformer-based coupled ocean-atmosphere model for ENSO studies. Sci. Bull. 69, 2323–2327 (2024).
Zhou, L. & Zhang, R.-H. A self-attention-based neural network for three-dimensional multivariate modeling and its skillful ENSO predictions. Sci. Adv. 9, eadf2827 (2023).
Qin, B. et al. The first kind of predictability problem of El Niño predictions in a multivariate coupled data-driven model. Quart. J. R. Meteorol. Soc. 150, 5452–5471 (2024).
Mu, B., Cui, Y., Yuan, S. & Qin, B. Incorporating heat budget dynamics in a Transformer-based deep learning model for skillful ENSO prediction. npj Clim. Atmos. Sci. 7, 208 (2024).
Sun, M., Chen, L., Li, T. & Luo, J.-J. CNN-based ENSO forecasts with a focus on SSTA zonal pattern and physical interpretation. Geophys Res Lett. 50, e2023GL105175 (2023).
Camps-Valls, G. et al. Artificial intelligence for modeling and understanding extreme weather and climate events. Nat. Commun. 16, 1919 (2025).
Li, Y., Zhang, Y. & Zheng, H. Evaluation of the Pangu model in forecasting rapid intensification of tropical cyclones. Atmos. Oceanic Sci. Lett. 100721 (2025).
Duan, S., Zhang, J., Bonfils, C. & Pallotta, G. Testing NeuralGCM’s capability to simulate future heatwaves based on the 2021 Pacific Northwest heatwave event. npj Clim. Atmos. Sci. 8, 251 (2025).
Demaeyer, J., Penny, S. G. & Vannitsem, S. Identifying Efficient Ensemble Perturbations for Initializing Subseasonal-To-Seasonal Prediction. J. Adv. Model. Earth Syst. 14, e2021MS002828 (2022).
Feng, J., Toth, Z., Zhang, J. & Peña, M. Ensemble forecasting: A foray of dynamics into the realm of statistics. Q. J. R. Meteorological Soc. 150, 2537–2560 (2024).
Wang, Y. et al. An ensemble-based coupled reanalysis of the climate from 1860 to the present (CoRea1860+). Earth Syst. Sci. Data Discuss. 1−38 (2025).
Fang, X. & Chen, N. Quantifying the Predictability of ENSO Complexity Using a Statistically Accurate Multiscale Stochastic Model and Information Theory. J. Clim. 36, 2681–2702 (2023).
Yeager, S. G. et al. The Seasonal-to-Multiyear Large Ensemble (SMYLE) prediction system using the Community Earth System Model version 2. Geosci. Model Dev. 15, 6451–6493 (2022).
Schevenhoven, F. et al. Supermodeling: Improving Predictions with an Ensemble of Interacting Models. Bull. Am. Meteorol. Soc. 104, E1670–E1686 (2023).
Larson, S. M. & Kirtman, B. P. Linking preconditioning to extreme ENSO events and reduced ensemble spread. Clim. Dyn. 52, 7417–7433 (2017).
Lorenz, E. N. Deterministic Nonperiodic Flow. J. Atmos. Sci. 20, 130–141 (1963).
Selz, T. & Craig, G. C. Can Artificial Intelligence-Based Weather Prediction Models Simulate the Butterfly Effect? Geophys. Res. Lett. 50, e2023GL105747 (2023).
Duan, W. & Huo, Z. An Approach to Generating Mutually Independent Initial Perturbations for Ensemble Forecasts: Orthogonal Conditional Nonlinear Optimal Perturbations. J. Atmos. Sci. 73, 997–1014 (2016).
Mu, M. & Duan, W. A new approach to studying ENSO predictability: Conditional nonlinear optimal perturbation. Chin. Sci. Bull. 48, 1045–1047 (2003).
Huo, Z. & Duan, W. The application of the orthogonal conditional nonlinear optimal perturbations method to typhoon track ensemble forecasts. Sci. China.: Earth Sci. 62, 376–388 (2019).
Huo, Z., Duan, W. & Zhou, F. Ensemble forecasts of tropical cyclone track with orthogonal conditional nonlinear optimal perturbations. Adv. Atmos. Sci. 36, 231–247 (2019).
Duan, W., Ma, J. & Vannitsem, S. An Ensemble Forecasting Method for Dealing with the Combined Effects of the Initial and Model Errors and a Potential Deep Learning Implementation. Mon. Weather Rev. 150, 2959–2976 (2022).
Zhang, H., Duan, W. & Zhang, Y. Using the orthogonal conditional nonlinear optimal perturbations approach to address the uncertainties of tropical cyclone track forecasts generated by the WRF model. Weather Forecast. 38, 1907–1933 (2023).
Zhou, L. & Zhang, R.-H. The 3D-Geoformer for ENSO studies: a Transformer-based model with integrated gradient methods for enhanced explainability. J. Oceanol. Limnol. 43, 1688–1708 (2025).
Gao, C., Zhou, L. & Zhang, R.-H. A Transformer-Based Deep Learning Model for Successful Predictions of the 2021 Second-Year La Niña Condition. Geophys. Res. Lett. 50, e2023GL104034 (2023).
Zhang, R.-H., Zhou, L., Gao, C. & Tao, L. Real-time predictions of the 2023-2024 climate conditions in the tropical Pacific using a purely data-driven Transformer model. Sci. China: Earth Sci. 67, 3709–3726 (2024).
Zhou, L. & Zhang, R.-H. ENSO-Related Precursor Pathways of Interannual Thermal Anomalies Identified Using a Transformer-Based Deep Learning Model in the Tropical Pacific. Geophys. Res. Lett. 51, e2023GL107347 (2024).
Chen, Y. et al. Combined dynamical-deep learning ENSO forecasts. Nat. Commun. 16, 3845 (2025).
Meng, Z. & Hakim, G. J. Reconstructing the tropical Pacific upper ocean using online data assimilation with a deep learning model. J. Adv. Modeling Earth Syst. 16, e2024MS004422 (2024).
Toth, Z. & Kalnay, E. Ensemble forecasting at NMC: The generation of perturbations. Bull. Am. Meteorol. Soc. 74, 2317–2330 (1993).
Duan, W., Yang, L., Xu, Z. & Chen, J. Conditional nonlinear optimal perturbation: Applications to ensemble forecasting of high-impact weather systems. In: Numerical Weather Prediction: East Asian Perspectives. Springer (2023).
Mu, M. & Jiang, Z. A new approach to the generation of initial perturbations for ensemble prediction: Conditional nonlinear optimal perturbation. Chin. Sci. Bull. 53, 2062–2068 (2008).
Pu, J., Mu, M., Feng, J., Zhong, X. & Li, H. A fast physics-based perturbation generator of machine learning weather model for efficient ensemble forecasts of tropical cyclone track. npj Clim. Atmos. Sci. 8, 128 (2025).
Duan, W. & Hu, J. The initial errors that induce a significant “spring predictability barrier” for El Niño events and their implications for target observation: results from an earth system model. Clim. Dyn. 46, 3599–3615 (2015).
Duan, W. & Mu, M. Predictability of El Niño-southern oscillation events. In: Oxford Research Encyclopedia of Climate Science) (2018).
Tao, L. et al. Impacts of Initial Zonal Current Errors on the Predictions of Two Types of El Niño Events. J. Geophys. Res. Oceans. 128, e2023JC019833 (2023).
Mlakar, P., Merse, J. & Faganeli Pucer, J. Ensemble weather forecast post-processing with a flexible probabilistic neural network approach. Q. J. R. Meteorological Soc. 150, 4156–4177 (2024).
Tziperman, E., Stone, L., Cane, M. A. & Jarosh, H. El Niño chaos: Overlapping of resonances between the seasonal cycle and the Pacific ocean-atmosphere oscillator. Science 264, 72–74 (1994).
Ludescher, J. et al. Improved El Niño forecasting by cooperativity detection. Proc. Natl. Acad. Sci. 110, 11742–11745 (2013).
Ludescher, J. et al. Very early warning of next El Niño. Proc. Natl. Acad. Sci. 111, 2064–2066 (2014).
Eyring, V. et al. Pushing the frontiers in climate modelling and analysis with machine learning. Nat. Clim. Change 14, 916–928 (2024).
Anderson, G. J. & Lucas, D. D. Machine Learning Predictions of a Multiresolution Climate Model Ensemble. Geophys Res Lett. 45, 4273–4280 (2018).
Zhang, Z., Fischer, E., Zscheischler, J. & Engelke, S. Numerical models outperform AI weather forecasts of record-breaking extremes. arXiv:250815724 (2025).
Charlton-Perez, A. J. et al. Do AI models produce better weather forecasts than physics-based models? A quantitative evaluation case study of Storm Ciarán. npj Clim. Atmos. Sci. 7, (2024).
Takahashi, K., Montecinos, A., Goubanova, K. & Dewitte, B. ENSO regimes: Reinterpreting the canonical and Modoki El Niño. Geophys. Res. Lett. 38, L10704 (2011).
Pan, B. et al. Improving Seasonal Forecast Using Probabilistic Deep Learning. J. Adv. Model. Earth Syst. 14, e2021MS002766 (2022).
Li, L., Carver, R., Lopez-Gomez, I., Sha, F. & Anderson, J. Generative emulation of weather forecast ensembles with diffusion models. Sci. Adv. 10, eadk4489 (2024).
Weyn, J. A., Durran, D. R., Caruana, R. & Cresswell-Clay, N. Sub-Seasonal Forecasting With a Large Ensemble of Deep-Learning Weather Prediction Models. J. Adv. Model. Earth Syst. 13, e2021MS002502 (2021).
Behringer, D. & Xue, Y. Evaluation of the global ocean data assimilation system at NCEP: The Pacific Ocean. In: Eighth Symp. on Integrated Observing and Assimilation Systems for Atmosphere, Oceans, and Land Surface) (2004).
Acknowledgements
This work is supported by the National Natural Science Foundation of China (No. 42030410), Laoshan Laboratory (No. LSKJ202202402), National Key R&D Program of China (No. 2024YFC2815702) and Jiangsu Innovation Research Group (No. JSSCTD 202346). Zhou is additionally supported by the National Natural Science Foundation of China (No. 42506019), Basic Research Program of Jiangsu Province (No. BK20250752), China National Postdoctoral Program for Innovative Talents (No. BX20240169) and China Postdoctoral Science Foundation (No. 2141062400101).
Author information
Authors and Affiliations
Contributions
Z.L., Z.R.-H. and T.J.L. designed the research. Z.L. performed the analysis and plotted figures. Z.L. and Z.R.-H. drafted the manuscript. All the authors contributed to physical interpretation and manuscript revision.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Zhou, L., Zhang, RH. & Tao, L. AI-Enabled conditional nonlinear optimal perturbation enhances ensemble prediction of extreme El Niño events. npj Clim Atmos Sci (2025). https://doi.org/10.1038/s41612-025-01303-6
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41612-025-01303-6
