Correcting errors in seasonal Arctic sea ice prediction of Earth system models with machine learning

<p>While Earth system models are essential for seasonal Arctic sea ice prediction, they often exhibit significant errors that are challenging to correct. In this study, we integrate a multilayer perceptron (MLP) machine learning (ML) model into the Norwegian Climate Prediction Model (NorCPM) t...

Full description

Saved in:
Bibliographic Details
Main Authors: Z. He, Y. Wang, J. Brajard, X. Wang, Z. Shen
Format: Article
Language:English
Published: Copernicus Publications 2025-08-01
Series:The Cryosphere
Online Access:https://tc.copernicus.org/articles/19/3279/2025/tc-19-3279-2025.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849222414871822336
author Z. He
Z. He
Z. He
Y. Wang
J. Brajard
X. Wang
X. Wang
Z. Shen
Z. Shen
author_facet Z. He
Z. He
Z. He
Y. Wang
J. Brajard
X. Wang
X. Wang
Z. Shen
Z. Shen
author_sort Z. He
collection DOAJ
description <p>While Earth system models are essential for seasonal Arctic sea ice prediction, they often exhibit significant errors that are challenging to correct. In this study, we integrate a multilayer perceptron (MLP) machine learning (ML) model into the Norwegian Climate Prediction Model (NorCPM) to improve seasonal sea ice predictions. We compare the online and offline error correction approaches. In the online approach, ML corrects errors in the model's instantaneous state during the model simulation, while in the offline approach, ML post-processes and calibrates predictions after the model simulation. Our results show that the ML models effectively learn and correct dynamical model errors in both approaches, leading to improved predictions of Arctic sea ice during the test period (i.e., 2003–2021). Both approaches yield the most significant improvements in the marginal ice zone, where error reductions in sea ice concentration exceed <span class="inline-formula">20 %</span>. These improvements vary seasonally, with the most substantial enhancements occurring in the Atlantic, Siberian, and Pacific regions from September to January. The offline error correction approach consistently outperforms the online error correction approach. This is primarily because the online approach targets only instantaneous model errors on the 15th of each month, while errors can grow during the subsequent 1-month model integration due to interactions among the model components, damping the error correction in monthly averages. Notably, in September, the online approach reduces the error of the pan-Arctic sea ice extent by <span class="inline-formula">50 %</span>, while the offline approach achieves a <span class="inline-formula">75 %</span> error reduction.</p>
format Article
id doaj-art-a259a0f134b34d3586a20c2dee4cc6d8
institution Kabale University
issn 1994-0416
1994-0424
language English
publishDate 2025-08-01
publisher Copernicus Publications
record_format Article
series The Cryosphere
spelling doaj-art-a259a0f134b34d3586a20c2dee4cc6d82025-08-26T05:30:44ZengCopernicus PublicationsThe Cryosphere1994-04161994-04242025-08-01193279329310.5194/tc-19-3279-2025Correcting errors in seasonal Arctic sea ice prediction of Earth system models with machine learningZ. He0Z. He1Z. He2Y. Wang3J. Brajard4X. Wang5X. Wang6Z. Shen7Z. Shen8Key Laboratory of Marine Hazards Forecasting, Ministry of Natural Resources, Hohai University, Nanjing, 210098, ChinaCollege of Oceanography, Hohai University, Nanjing, 210098, ChinaNansen Environmental and Remote Sensing Center and Bjerknes Centre for Climate Research, Jahnebakken 3, Bergen, 5007, NorwayNansen Environmental and Remote Sensing Center and Bjerknes Centre for Climate Research, Jahnebakken 3, Bergen, 5007, NorwayNansen Environmental and Remote Sensing Center and Bjerknes Centre for Climate Research, Jahnebakken 3, Bergen, 5007, NorwayKey Laboratory of Marine Hazards Forecasting, Ministry of Natural Resources, Hohai University, Nanjing, 210098, ChinaCollege of Oceanography, Hohai University, Nanjing, 210098, ChinaKey Laboratory of Marine Hazards Forecasting, Ministry of Natural Resources, Hohai University, Nanjing, 210098, ChinaCollege of Oceanography, Hohai University, Nanjing, 210098, China<p>While Earth system models are essential for seasonal Arctic sea ice prediction, they often exhibit significant errors that are challenging to correct. In this study, we integrate a multilayer perceptron (MLP) machine learning (ML) model into the Norwegian Climate Prediction Model (NorCPM) to improve seasonal sea ice predictions. We compare the online and offline error correction approaches. In the online approach, ML corrects errors in the model's instantaneous state during the model simulation, while in the offline approach, ML post-processes and calibrates predictions after the model simulation. Our results show that the ML models effectively learn and correct dynamical model errors in both approaches, leading to improved predictions of Arctic sea ice during the test period (i.e., 2003–2021). Both approaches yield the most significant improvements in the marginal ice zone, where error reductions in sea ice concentration exceed <span class="inline-formula">20 %</span>. These improvements vary seasonally, with the most substantial enhancements occurring in the Atlantic, Siberian, and Pacific regions from September to January. The offline error correction approach consistently outperforms the online error correction approach. This is primarily because the online approach targets only instantaneous model errors on the 15th of each month, while errors can grow during the subsequent 1-month model integration due to interactions among the model components, damping the error correction in monthly averages. Notably, in September, the online approach reduces the error of the pan-Arctic sea ice extent by <span class="inline-formula">50 %</span>, while the offline approach achieves a <span class="inline-formula">75 %</span> error reduction.</p>https://tc.copernicus.org/articles/19/3279/2025/tc-19-3279-2025.pdf
spellingShingle Z. He
Z. He
Z. He
Y. Wang
J. Brajard
X. Wang
X. Wang
Z. Shen
Z. Shen
Correcting errors in seasonal Arctic sea ice prediction of Earth system models with machine learning
The Cryosphere
title Correcting errors in seasonal Arctic sea ice prediction of Earth system models with machine learning
title_full Correcting errors in seasonal Arctic sea ice prediction of Earth system models with machine learning
title_fullStr Correcting errors in seasonal Arctic sea ice prediction of Earth system models with machine learning
title_full_unstemmed Correcting errors in seasonal Arctic sea ice prediction of Earth system models with machine learning
title_short Correcting errors in seasonal Arctic sea ice prediction of Earth system models with machine learning
title_sort correcting errors in seasonal arctic sea ice prediction of earth system models with machine learning
url https://tc.copernicus.org/articles/19/3279/2025/tc-19-3279-2025.pdf
work_keys_str_mv AT zhe correctingerrorsinseasonalarcticseaicepredictionofearthsystemmodelswithmachinelearning
AT zhe correctingerrorsinseasonalarcticseaicepredictionofearthsystemmodelswithmachinelearning
AT zhe correctingerrorsinseasonalarcticseaicepredictionofearthsystemmodelswithmachinelearning
AT ywang correctingerrorsinseasonalarcticseaicepredictionofearthsystemmodelswithmachinelearning
AT jbrajard correctingerrorsinseasonalarcticseaicepredictionofearthsystemmodelswithmachinelearning
AT xwang correctingerrorsinseasonalarcticseaicepredictionofearthsystemmodelswithmachinelearning
AT xwang correctingerrorsinseasonalarcticseaicepredictionofearthsystemmodelswithmachinelearning
AT zshen correctingerrorsinseasonalarcticseaicepredictionofearthsystemmodelswithmachinelearning
AT zshen correctingerrorsinseasonalarcticseaicepredictionofearthsystemmodelswithmachinelearning