Assessment of High‐Resolution Dynamical and Machine Learning Models for Prediction of Sea Ice Concentration in a Regional Application
In this study, the potential for sea ice concentration prediction using machine‐learning methods is investigated. Three different sea ice prediction models are compared: one high‐resolution dynamical assimilative model and two statistical machine‐learning models. The properties of all three models a...
Published in: | Journal of Geophysical Research: Oceans |
---|---|
Main Authors: | , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
American Geophysical Union (AGU)
2020
|
Subjects: | |
Online Access: | https://hdl.handle.net/10037/19736 https://doi.org/10.1029/2020JC016277 |
id |
ftunivtroemsoe:oai:munin.uit.no:10037/19736 |
---|---|
record_format |
openpolar |
spelling |
ftunivtroemsoe:oai:munin.uit.no:10037/19736 2023-07-16T03:55:33+02:00 Assessment of High‐Resolution Dynamical and Machine Learning Models for Prediction of Sea Ice Concentration in a Regional Application Fritzner, Sindre Markus Graversen, Rune Christensen, Kai Håkon 2020-10-17 https://hdl.handle.net/10037/19736 https://doi.org/10.1029/2020JC016277 eng eng American Geophysical Union (AGU) Journal of Geophysical Research (JGR): Oceans info:eu-repo/grantAgreement/RCN/SFI/237906/Norway/Centre for Integrated Remote Sensing and Forecasting for Arctic Operations/CIRFA/ Fritzner SM, Graversen R, Christensen KH. Assessment of High‐Resolution Dynamical and Machine Learning Models for Prediction of Sea Ice Concentration in a Regional Application. Journal of Geophysical Research (JGR): Oceans. 2020;125(11):1-23 FRIDAID 1843517 doi:10.1029/2020JC016277 2169-9275 2169-9291 https://hdl.handle.net/10037/19736 openAccess ©2020. American Geophysical Union. All Rights Reserved. VDP::Mathematics and natural science: 400::Physics: 430 VDP::Matematikk og Naturvitenskap: 400::Fysikk: 430 Journal article Tidsskriftartikkel submittedVersion 2020 ftunivtroemsoe https://doi.org/10.1029/2020JC016277 2023-06-28T23:06:34Z In this study, the potential for sea ice concentration prediction using machine‐learning methods is investigated. Three different sea ice prediction models are compared: one high‐resolution dynamical assimilative model and two statistical machine‐learning models. The properties of all three models are explored, and the quality of their forecasts is compared. The dynamical model is a state‐of‐the‐art coupled ocean and sea ice ensemble‐prediction system with assimilation. The observations assimilated are high‐resolution sea ice concentration from synthetic aperture radar (SAR) and sea surface temperature from infrared instruments. The machine‐learning prediction models are a fully convolutional network and a k ‐nearest neighbors method. These methods use several variables as input for the prediction: sea ice concentration, sea surface temperature, and 2‐m air temperature. Earlier studies have applied machine‐learning approaches primarily for seasonal ice forecast. Here we focus on short‐term predictions with a length of 1–4 weeks, which are of high interest for marine operations. The goal is to predict the future state of the sea ice using the same categories as traditional ice charts. The machine‐learning forecasts were compared to persistence, which is the assumption that the sea ice does not change over the forecasting period. The machine‐learning forecasts were found to improve upon persistence in periods of substantial change. In addition, compared to the dynamical model, the k ‐nearest neighbor algorithm was found to improve upon the 7‐day forecast during a period of small sea ice variations. The fully convolutional network provided similar quality as the dynamical forecast. The study shows that there is a potential for sea ice predictions using machine‐learning methods. Article in Journal/Newspaper Arctic Sea ice University of Tromsø: Munin Open Research Archive Journal of Geophysical Research: Oceans 125 11 |
institution |
Open Polar |
collection |
University of Tromsø: Munin Open Research Archive |
op_collection_id |
ftunivtroemsoe |
language |
English |
topic |
VDP::Mathematics and natural science: 400::Physics: 430 VDP::Matematikk og Naturvitenskap: 400::Fysikk: 430 |
spellingShingle |
VDP::Mathematics and natural science: 400::Physics: 430 VDP::Matematikk og Naturvitenskap: 400::Fysikk: 430 Fritzner, Sindre Markus Graversen, Rune Christensen, Kai Håkon Assessment of High‐Resolution Dynamical and Machine Learning Models for Prediction of Sea Ice Concentration in a Regional Application |
topic_facet |
VDP::Mathematics and natural science: 400::Physics: 430 VDP::Matematikk og Naturvitenskap: 400::Fysikk: 430 |
description |
In this study, the potential for sea ice concentration prediction using machine‐learning methods is investigated. Three different sea ice prediction models are compared: one high‐resolution dynamical assimilative model and two statistical machine‐learning models. The properties of all three models are explored, and the quality of their forecasts is compared. The dynamical model is a state‐of‐the‐art coupled ocean and sea ice ensemble‐prediction system with assimilation. The observations assimilated are high‐resolution sea ice concentration from synthetic aperture radar (SAR) and sea surface temperature from infrared instruments. The machine‐learning prediction models are a fully convolutional network and a k ‐nearest neighbors method. These methods use several variables as input for the prediction: sea ice concentration, sea surface temperature, and 2‐m air temperature. Earlier studies have applied machine‐learning approaches primarily for seasonal ice forecast. Here we focus on short‐term predictions with a length of 1–4 weeks, which are of high interest for marine operations. The goal is to predict the future state of the sea ice using the same categories as traditional ice charts. The machine‐learning forecasts were compared to persistence, which is the assumption that the sea ice does not change over the forecasting period. The machine‐learning forecasts were found to improve upon persistence in periods of substantial change. In addition, compared to the dynamical model, the k ‐nearest neighbor algorithm was found to improve upon the 7‐day forecast during a period of small sea ice variations. The fully convolutional network provided similar quality as the dynamical forecast. The study shows that there is a potential for sea ice predictions using machine‐learning methods. |
format |
Article in Journal/Newspaper |
author |
Fritzner, Sindre Markus Graversen, Rune Christensen, Kai Håkon |
author_facet |
Fritzner, Sindre Markus Graversen, Rune Christensen, Kai Håkon |
author_sort |
Fritzner, Sindre Markus |
title |
Assessment of High‐Resolution Dynamical and Machine Learning Models for Prediction of Sea Ice Concentration in a Regional Application |
title_short |
Assessment of High‐Resolution Dynamical and Machine Learning Models for Prediction of Sea Ice Concentration in a Regional Application |
title_full |
Assessment of High‐Resolution Dynamical and Machine Learning Models for Prediction of Sea Ice Concentration in a Regional Application |
title_fullStr |
Assessment of High‐Resolution Dynamical and Machine Learning Models for Prediction of Sea Ice Concentration in a Regional Application |
title_full_unstemmed |
Assessment of High‐Resolution Dynamical and Machine Learning Models for Prediction of Sea Ice Concentration in a Regional Application |
title_sort |
assessment of high‐resolution dynamical and machine learning models for prediction of sea ice concentration in a regional application |
publisher |
American Geophysical Union (AGU) |
publishDate |
2020 |
url |
https://hdl.handle.net/10037/19736 https://doi.org/10.1029/2020JC016277 |
genre |
Arctic Sea ice |
genre_facet |
Arctic Sea ice |
op_relation |
Journal of Geophysical Research (JGR): Oceans info:eu-repo/grantAgreement/RCN/SFI/237906/Norway/Centre for Integrated Remote Sensing and Forecasting for Arctic Operations/CIRFA/ Fritzner SM, Graversen R, Christensen KH. Assessment of High‐Resolution Dynamical and Machine Learning Models for Prediction of Sea Ice Concentration in a Regional Application. Journal of Geophysical Research (JGR): Oceans. 2020;125(11):1-23 FRIDAID 1843517 doi:10.1029/2020JC016277 2169-9275 2169-9291 https://hdl.handle.net/10037/19736 |
op_rights |
openAccess ©2020. American Geophysical Union. All Rights Reserved. |
op_doi |
https://doi.org/10.1029/2020JC016277 |
container_title |
Journal of Geophysical Research: Oceans |
container_volume |
125 |
container_issue |
11 |
_version_ |
1771541774100070400 |