Toward long-range ENSO prediction with an explainable deep learning model

Abstract El Niño-Southern Oscillation (ENSO) is a prominent mode of interannual climate variability with far-reaching global impacts. Its evolution is governed by intricate air-sea interactions, posing significant challenges for long-term prediction. In this study, we introduce CTEFNet, a multivaria...

Full description

Saved in:
Bibliographic Details
Main Authors: Qi Chen, Yinghao Cui, Guobin Hong, Karumuri Ashok, Yuchun Pu, Xiaogu Zheng, Xuanze Zhang, Wei Zhong, Peng Zhan, Zhonglei Wang
Format: Article
Language:English
Published: Nature Portfolio 2025-07-01
Series:npj Climate and Atmospheric Science
Online Access:https://doi.org/10.1038/s41612-025-01159-w
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849237679205515264
author Qi Chen
Yinghao Cui
Guobin Hong
Karumuri Ashok
Yuchun Pu
Xiaogu Zheng
Xuanze Zhang
Wei Zhong
Peng Zhan
Zhonglei Wang
author_facet Qi Chen
Yinghao Cui
Guobin Hong
Karumuri Ashok
Yuchun Pu
Xiaogu Zheng
Xuanze Zhang
Wei Zhong
Peng Zhan
Zhonglei Wang
author_sort Qi Chen
collection DOAJ
description Abstract El Niño-Southern Oscillation (ENSO) is a prominent mode of interannual climate variability with far-reaching global impacts. Its evolution is governed by intricate air-sea interactions, posing significant challenges for long-term prediction. In this study, we introduce CTEFNet, a multivariate deep learning model that synergizes convolutional neural networks and transformers to enhance ENSO forecasting. By integrating multiple oceanic and atmospheric predictors, CTEFNet extends the effective forecast lead time to 20 months while mitigating the impact of the spring predictability barrier, outperforming both dynamical models and state-of-the-art deep learning approaches. Furthermore, CTEFNet offers physically meaningful and statistically significant insights through gradient-based sensitivity analysis, revealing the key precursor signals that govern ENSO dynamics, which align with well-established theories and reveal new insights about inter-basin interactions among the Pacific, Atlantic, and Indian Oceans. The CTEFNet’s superior predictive skill and interpretable sensitivity assessments underscore its potential for advancing climate prediction. Our findings highlight the importance of multivariate coupling in ENSO evolution and demonstrate the promise of deep learning in capturing complex climate dynamics with enhanced interpretability.
format Article
id doaj-art-38f29f6eaedd4832ae8324ee28e68b03
institution Kabale University
issn 2397-3722
language English
publishDate 2025-07-01
publisher Nature Portfolio
record_format Article
series npj Climate and Atmospheric Science
spelling doaj-art-38f29f6eaedd4832ae8324ee28e68b032025-08-20T04:01:53ZengNature Portfolionpj Climate and Atmospheric Science2397-37222025-07-01811910.1038/s41612-025-01159-wToward long-range ENSO prediction with an explainable deep learning modelQi Chen0Yinghao Cui1Guobin Hong2Karumuri Ashok3Yuchun Pu4Xiaogu Zheng5Xuanze Zhang6Wei Zhong7Peng Zhan8Zhonglei Wang9Department of Ocean Science and Engineering, Southern University of Science and TechnologyDepartment of Statistics and Data Science, School of Economic, Xiamen UniversityMOE Key Laboratory of Econometrics, Xiamen UniversityCentre for Earth, Ocean and Atmospheric Sciences, University of HyderabadMeituanShanghai Zhangjiang Institute of MathematicsKey Laboratory of Water Cycle and Related Land Surface Processes, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of SciencesDepartment of Statistics and Data Science, School of Economic, Xiamen UniversityDepartment of Ocean Science and Engineering, Southern University of Science and TechnologyDepartment of Statistics and Data Science, School of Economic, Xiamen UniversityAbstract El Niño-Southern Oscillation (ENSO) is a prominent mode of interannual climate variability with far-reaching global impacts. Its evolution is governed by intricate air-sea interactions, posing significant challenges for long-term prediction. In this study, we introduce CTEFNet, a multivariate deep learning model that synergizes convolutional neural networks and transformers to enhance ENSO forecasting. By integrating multiple oceanic and atmospheric predictors, CTEFNet extends the effective forecast lead time to 20 months while mitigating the impact of the spring predictability barrier, outperforming both dynamical models and state-of-the-art deep learning approaches. Furthermore, CTEFNet offers physically meaningful and statistically significant insights through gradient-based sensitivity analysis, revealing the key precursor signals that govern ENSO dynamics, which align with well-established theories and reveal new insights about inter-basin interactions among the Pacific, Atlantic, and Indian Oceans. The CTEFNet’s superior predictive skill and interpretable sensitivity assessments underscore its potential for advancing climate prediction. Our findings highlight the importance of multivariate coupling in ENSO evolution and demonstrate the promise of deep learning in capturing complex climate dynamics with enhanced interpretability.https://doi.org/10.1038/s41612-025-01159-w
spellingShingle Qi Chen
Yinghao Cui
Guobin Hong
Karumuri Ashok
Yuchun Pu
Xiaogu Zheng
Xuanze Zhang
Wei Zhong
Peng Zhan
Zhonglei Wang
Toward long-range ENSO prediction with an explainable deep learning model
npj Climate and Atmospheric Science
title Toward long-range ENSO prediction with an explainable deep learning model
title_full Toward long-range ENSO prediction with an explainable deep learning model
title_fullStr Toward long-range ENSO prediction with an explainable deep learning model
title_full_unstemmed Toward long-range ENSO prediction with an explainable deep learning model
title_short Toward long-range ENSO prediction with an explainable deep learning model
title_sort toward long range enso prediction with an explainable deep learning model
url https://doi.org/10.1038/s41612-025-01159-w
work_keys_str_mv AT qichen towardlongrangeensopredictionwithanexplainabledeeplearningmodel
AT yinghaocui towardlongrangeensopredictionwithanexplainabledeeplearningmodel
AT guobinhong towardlongrangeensopredictionwithanexplainabledeeplearningmodel
AT karumuriashok towardlongrangeensopredictionwithanexplainabledeeplearningmodel
AT yuchunpu towardlongrangeensopredictionwithanexplainabledeeplearningmodel
AT xiaoguzheng towardlongrangeensopredictionwithanexplainabledeeplearningmodel
AT xuanzezhang towardlongrangeensopredictionwithanexplainabledeeplearningmodel
AT weizhong towardlongrangeensopredictionwithanexplainabledeeplearningmodel
AT pengzhan towardlongrangeensopredictionwithanexplainabledeeplearningmodel
AT zhongleiwang towardlongrangeensopredictionwithanexplainabledeeplearningmodel