ENSO‐Related Precursor Pathways of Interannual Thermal Anomalies Identified Using a Transformer‐Based Deep Learning Model in the Tropical Pacific

Abstract Recent studies have demonstrated great values of deep‐learning (DL) methods for improving El Niño‐Southern Oscillation (ENSO) predictions. However, the black‐box nature of DL makes it challenging to physically interpret mechanisms responsible for successful ENSO predictions. Here, we demons...

Full description

Saved in:
Bibliographic Details
Main Authors: Lu Zhou, Rong‐Hua Zhang
Format: Article
Language:English
Published: Wiley 2024-06-01
Series:Geophysical Research Letters
Subjects:
Online Access:https://doi.org/10.1029/2023GL107347
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850134804093206528
author Lu Zhou
Rong‐Hua Zhang
author_facet Lu Zhou
Rong‐Hua Zhang
author_sort Lu Zhou
collection DOAJ
description Abstract Recent studies have demonstrated great values of deep‐learning (DL) methods for improving El Niño‐Southern Oscillation (ENSO) predictions. However, the black‐box nature of DL makes it challenging to physically interpret mechanisms responsible for successful ENSO predictions. Here, we demonstrate an interpretable method by performing perturbation experiments to predictors and quantifying input‐output relationships in predictions by using a transformer‐based model; ENSO‐related thermal precursors serving as initial conditions during multi‐month time intervals (TIs) are identified in the equatorial‐northern Pacific, acting to precondition input predictors to provide for long‐lead ENSO predictability. Results reveal the existence of upper‐ocean temperature anomaly pathways and consistent phase propagations of thermal precursors around the tropical Pacific. It is illustrated that three‐dimensional thermal fields and their basinwide evolution during long TIs act to enhance long‐lead prediction skills of ENSO. These physically explainable results indicate that neural networks can adequately represent predictable precursors in the input predictors for successful ENSO predictions.
format Article
id doaj-art-017ba7839d034150bb488f5751d13aed
institution OA Journals
issn 0094-8276
1944-8007
language English
publishDate 2024-06-01
publisher Wiley
record_format Article
series Geophysical Research Letters
spelling doaj-art-017ba7839d034150bb488f5751d13aed2025-08-20T02:31:38ZengWileyGeophysical Research Letters0094-82761944-80072024-06-015112n/an/a10.1029/2023GL107347ENSO‐Related Precursor Pathways of Interannual Thermal Anomalies Identified Using a Transformer‐Based Deep Learning Model in the Tropical PacificLu Zhou0Rong‐Hua Zhang1Key Laboratory of Ocean Circulation and Waves Institute of Oceanology Chinese Academy of Sciences Qingdao ChinaUniversity of Chinese Academy of Sciences Beijing ChinaAbstract Recent studies have demonstrated great values of deep‐learning (DL) methods for improving El Niño‐Southern Oscillation (ENSO) predictions. However, the black‐box nature of DL makes it challenging to physically interpret mechanisms responsible for successful ENSO predictions. Here, we demonstrate an interpretable method by performing perturbation experiments to predictors and quantifying input‐output relationships in predictions by using a transformer‐based model; ENSO‐related thermal precursors serving as initial conditions during multi‐month time intervals (TIs) are identified in the equatorial‐northern Pacific, acting to precondition input predictors to provide for long‐lead ENSO predictability. Results reveal the existence of upper‐ocean temperature anomaly pathways and consistent phase propagations of thermal precursors around the tropical Pacific. It is illustrated that three‐dimensional thermal fields and their basinwide evolution during long TIs act to enhance long‐lead prediction skills of ENSO. These physically explainable results indicate that neural networks can adequately represent predictable precursors in the input predictors for successful ENSO predictions.https://doi.org/10.1029/2023GL107347ENSO predictionsthermal precursorsmultivariate three‐dimensional (3D) predictionstransformer‐based modelexplainable artificial intelligence (XAI)
spellingShingle Lu Zhou
Rong‐Hua Zhang
ENSO‐Related Precursor Pathways of Interannual Thermal Anomalies Identified Using a Transformer‐Based Deep Learning Model in the Tropical Pacific
Geophysical Research Letters
ENSO predictions
thermal precursors
multivariate three‐dimensional (3D) predictions
transformer‐based model
explainable artificial intelligence (XAI)
title ENSO‐Related Precursor Pathways of Interannual Thermal Anomalies Identified Using a Transformer‐Based Deep Learning Model in the Tropical Pacific
title_full ENSO‐Related Precursor Pathways of Interannual Thermal Anomalies Identified Using a Transformer‐Based Deep Learning Model in the Tropical Pacific
title_fullStr ENSO‐Related Precursor Pathways of Interannual Thermal Anomalies Identified Using a Transformer‐Based Deep Learning Model in the Tropical Pacific
title_full_unstemmed ENSO‐Related Precursor Pathways of Interannual Thermal Anomalies Identified Using a Transformer‐Based Deep Learning Model in the Tropical Pacific
title_short ENSO‐Related Precursor Pathways of Interannual Thermal Anomalies Identified Using a Transformer‐Based Deep Learning Model in the Tropical Pacific
title_sort enso related precursor pathways of interannual thermal anomalies identified using a transformer based deep learning model in the tropical pacific
topic ENSO predictions
thermal precursors
multivariate three‐dimensional (3D) predictions
transformer‐based model
explainable artificial intelligence (XAI)
url https://doi.org/10.1029/2023GL107347
work_keys_str_mv AT luzhou ensorelatedprecursorpathwaysofinterannualthermalanomaliesidentifiedusingatransformerbaseddeeplearningmodelinthetropicalpacific
AT ronghuazhang ensorelatedprecursorpathwaysofinterannualthermalanomaliesidentifiedusingatransformerbaseddeeplearningmodelinthetropicalpacific