Projection of ENSO using observation-informed deep learning

Abstract The El Niño-Southern Oscillation (ENSO) profoundly impacts global climate, but its sea surface temperature (SST) variability projected by climate models remains uncertain, with a substantial inter-model spread in 21st-century projections. Model-observation discrepancies in ENSO physics cont...

Full description

Saved in:
Bibliographic Details
Main Authors: Yuchao Zhu, Rong-Hua Zhang, Fan Wang, Wenju Cai, Delei Li, Shoude Guan, Yuanlong Li
Format: Article
Language:English
Published: Nature Portfolio 2025-08-01
Series:Nature Communications
Online Access:https://doi.org/10.1038/s41467-025-63157-z
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849226145308868608
author Yuchao Zhu
Rong-Hua Zhang
Fan Wang
Wenju Cai
Delei Li
Shoude Guan
Yuanlong Li
author_facet Yuchao Zhu
Rong-Hua Zhang
Fan Wang
Wenju Cai
Delei Li
Shoude Guan
Yuanlong Li
author_sort Yuchao Zhu
collection DOAJ
description Abstract The El Niño-Southern Oscillation (ENSO) profoundly impacts global climate, but its sea surface temperature (SST) variability projected by climate models remains uncertain, with a substantial inter-model spread in 21st-century projections. Model-observation discrepancies in ENSO physics contribute to this uncertainty, necessitating observational constraints to refine projections. However, methods to achieve this constraint remain unclear. Here, we show that deep learning informed by the observed response of ENSO SST variability to tropical Pacific warming patterns reduces projection uncertainty by 54% under a high-emission scenario. Specifically, artificial neural networks (ANNs), trained on climate model simulations and observations, successfully capture the real-world ENSO response. Interpretability analyses reveal that replicating observed ENSO physics by ANNs is critical, identifying warming in the far-eastern and central equatorial Pacific as key to ENSO change. A model-as-truth approach further confirms the robustness of ANN-generated projections. By conditioning future ENSO SST variability projection on the ANN-inferred ENSO response to tropical Pacific warming, uncertainty is reduced from a range of 0.59 °C to 0.27 °C. Our results highlight the prospect of integrating machine learning with observations to reduce uncertainty in climate projections.
format Article
id doaj-art-8b8e5da0636c4075865a42b28a2c0e86
institution Kabale University
issn 2041-1723
language English
publishDate 2025-08-01
publisher Nature Portfolio
record_format Article
series Nature Communications
spelling doaj-art-8b8e5da0636c4075865a42b28a2c0e862025-08-24T11:37:25ZengNature PortfolioNature Communications2041-17232025-08-0116111210.1038/s41467-025-63157-zProjection of ENSO using observation-informed deep learningYuchao Zhu0Rong-Hua Zhang1Fan Wang2Wenju Cai3Delei Li4Shoude Guan5Yuanlong Li6Key Laboratory of Ocean Observation and Forecasting & Laboratory of Ocean Circulation and Waves, Institute of Oceanology, Chinese Academy of SciencesState Key Laboratory of Climate System Prediction and Risk Management/School of Marine Sciences, Nanjing University of Information Science and TechnologyKey Laboratory of Ocean Observation and Forecasting & Laboratory of Ocean Circulation and Waves, Institute of Oceanology, Chinese Academy of SciencesFrontiers Science Center for Deep Ocean Multispheres and Earth System (FDOMES) and Key Laboratory of Physical Oceanography, Ocean University of ChinaLaoshan LaboratoryFrontiers Science Center for Deep Ocean Multispheres and Earth System (FDOMES) and Key Laboratory of Physical Oceanography, Ocean University of ChinaKey Laboratory of Ocean Observation and Forecasting & Laboratory of Ocean Circulation and Waves, Institute of Oceanology, Chinese Academy of SciencesAbstract The El Niño-Southern Oscillation (ENSO) profoundly impacts global climate, but its sea surface temperature (SST) variability projected by climate models remains uncertain, with a substantial inter-model spread in 21st-century projections. Model-observation discrepancies in ENSO physics contribute to this uncertainty, necessitating observational constraints to refine projections. However, methods to achieve this constraint remain unclear. Here, we show that deep learning informed by the observed response of ENSO SST variability to tropical Pacific warming patterns reduces projection uncertainty by 54% under a high-emission scenario. Specifically, artificial neural networks (ANNs), trained on climate model simulations and observations, successfully capture the real-world ENSO response. Interpretability analyses reveal that replicating observed ENSO physics by ANNs is critical, identifying warming in the far-eastern and central equatorial Pacific as key to ENSO change. A model-as-truth approach further confirms the robustness of ANN-generated projections. By conditioning future ENSO SST variability projection on the ANN-inferred ENSO response to tropical Pacific warming, uncertainty is reduced from a range of 0.59 °C to 0.27 °C. Our results highlight the prospect of integrating machine learning with observations to reduce uncertainty in climate projections.https://doi.org/10.1038/s41467-025-63157-z
spellingShingle Yuchao Zhu
Rong-Hua Zhang
Fan Wang
Wenju Cai
Delei Li
Shoude Guan
Yuanlong Li
Projection of ENSO using observation-informed deep learning
Nature Communications
title Projection of ENSO using observation-informed deep learning
title_full Projection of ENSO using observation-informed deep learning
title_fullStr Projection of ENSO using observation-informed deep learning
title_full_unstemmed Projection of ENSO using observation-informed deep learning
title_short Projection of ENSO using observation-informed deep learning
title_sort projection of enso using observation informed deep learning
url https://doi.org/10.1038/s41467-025-63157-z
work_keys_str_mv AT yuchaozhu projectionofensousingobservationinformeddeeplearning
AT ronghuazhang projectionofensousingobservationinformeddeeplearning
AT fanwang projectionofensousingobservationinformeddeeplearning
AT wenjucai projectionofensousingobservationinformeddeeplearning
AT deleili projectionofensousingobservationinformeddeeplearning
AT shoudeguan projectionofensousingobservationinformeddeeplearning
AT yuanlongli projectionofensousingobservationinformeddeeplearning