Convolutional Neural Networks Trained on Internal Variability Predict Forced Response of TOA Radiation by Learning the Pattern Effect

Abstract Predicting forced, long‐term radiative feedbacks from internal climate variability has been a decades‐long quest in climate science. We train a convolutional neural network (CNN) to predict annual‐ and global‐mean top of the atmosphere radiation anomalies from time‐varying maps of near‐surf...

Full description

Saved in:
Bibliographic Details
Main Authors: Maria Rugenstein, Senne VanLoon, Elizabeth A. Barnes
Format: Article
Language:English
Published: Wiley 2025-02-01
Series:Geophysical Research Letters
Online Access:https://doi.org/10.1029/2024GL109581
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849322175477055488
author Maria Rugenstein
Senne VanLoon
Elizabeth A. Barnes
author_facet Maria Rugenstein
Senne VanLoon
Elizabeth A. Barnes
author_sort Maria Rugenstein
collection DOAJ
description Abstract Predicting forced, long‐term radiative feedbacks from internal climate variability has been a decades‐long quest in climate science. We train a convolutional neural network (CNN) to predict annual‐ and global‐mean top of the atmosphere radiation anomalies from time‐varying maps of near‐surface temperature in climate models. Trained on internal variability alone, the nonlinear CNN can predict radiation under strong climate change, outperforms a regularized linear regression approach, and works within and across different climate models. We show with explainable artificial intelligence methods that the CNN draws predictive skill from physically meaningful regions but at much smaller spatial scales than currently assumed.
format Article
id doaj-art-0bd66aba94fd4cac8d906aeff6a8edaf
institution Kabale University
issn 0094-8276
1944-8007
language English
publishDate 2025-02-01
publisher Wiley
record_format Article
series Geophysical Research Letters
spelling doaj-art-0bd66aba94fd4cac8d906aeff6a8edaf2025-08-20T03:49:31ZengWileyGeophysical Research Letters0094-82761944-80072025-02-01524n/an/a10.1029/2024GL109581Convolutional Neural Networks Trained on Internal Variability Predict Forced Response of TOA Radiation by Learning the Pattern EffectMaria Rugenstein0Senne VanLoon1Elizabeth A. Barnes2Colorado State University Fort Collins CO USAColorado State University Fort Collins CO USAColorado State University Fort Collins CO USAAbstract Predicting forced, long‐term radiative feedbacks from internal climate variability has been a decades‐long quest in climate science. We train a convolutional neural network (CNN) to predict annual‐ and global‐mean top of the atmosphere radiation anomalies from time‐varying maps of near‐surface temperature in climate models. Trained on internal variability alone, the nonlinear CNN can predict radiation under strong climate change, outperforms a regularized linear regression approach, and works within and across different climate models. We show with explainable artificial intelligence methods that the CNN draws predictive skill from physically meaningful regions but at much smaller spatial scales than currently assumed.https://doi.org/10.1029/2024GL109581
spellingShingle Maria Rugenstein
Senne VanLoon
Elizabeth A. Barnes
Convolutional Neural Networks Trained on Internal Variability Predict Forced Response of TOA Radiation by Learning the Pattern Effect
Geophysical Research Letters
title Convolutional Neural Networks Trained on Internal Variability Predict Forced Response of TOA Radiation by Learning the Pattern Effect
title_full Convolutional Neural Networks Trained on Internal Variability Predict Forced Response of TOA Radiation by Learning the Pattern Effect
title_fullStr Convolutional Neural Networks Trained on Internal Variability Predict Forced Response of TOA Radiation by Learning the Pattern Effect
title_full_unstemmed Convolutional Neural Networks Trained on Internal Variability Predict Forced Response of TOA Radiation by Learning the Pattern Effect
title_short Convolutional Neural Networks Trained on Internal Variability Predict Forced Response of TOA Radiation by Learning the Pattern Effect
title_sort convolutional neural networks trained on internal variability predict forced response of toa radiation by learning the pattern effect
url https://doi.org/10.1029/2024GL109581
work_keys_str_mv AT mariarugenstein convolutionalneuralnetworkstrainedoninternalvariabilitypredictforcedresponseoftoaradiationbylearningthepatterneffect
AT sennevanloon convolutionalneuralnetworkstrainedoninternalvariabilitypredictforcedresponseoftoaradiationbylearningthepatterneffect
AT elizabethabarnes convolutionalneuralnetworkstrainedoninternalvariabilitypredictforcedresponseoftoaradiationbylearningthepatterneffect