Predicting ocean-induced ice-shelf melt rates using deep learning
Through their role in buttressing upstream ice flow, Antarctic ice shelves play an important part in regulating future sea-level change. Reduction in ice-shelf buttressing caused by increased ocean-induced melt along their undersides is now understood to be one of the key drivers of ice loss from th...
Published in: | The Cryosphere |
---|---|
Main Authors: | , , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
Coperincus
2023
|
Subjects: | |
Online Access: | https://nrl.northumbria.ac.uk/id/eprint/51332/ https://doi.org/10.5194/tc-17-499-2023 https://nrl.northumbria.ac.uk/id/eprint/51332/1/tc_17_499_2023.pdf |
id |
ftunivnorthumb:oai:nrl.northumbria.ac.uk:51332 |
---|---|
record_format |
openpolar |
spelling |
ftunivnorthumb:oai:nrl.northumbria.ac.uk:51332 2023-05-15T13:51:03+02:00 Predicting ocean-induced ice-shelf melt rates using deep learning Rosier, Sebastian Bull, Christopher Woo, Wai Lok Gudmundsson, Hilmar 2023-02-07 text https://nrl.northumbria.ac.uk/id/eprint/51332/ https://doi.org/10.5194/tc-17-499-2023 https://nrl.northumbria.ac.uk/id/eprint/51332/1/tc_17_499_2023.pdf en eng Coperincus https://nrl.northumbria.ac.uk/id/eprint/51332/1/tc_17_499_2023.pdf Rosier, Sebastian, Bull, Christopher, Woo, Wai Lok and Gudmundsson, Hilmar (2023) Predicting ocean-induced ice-shelf melt rates using deep learning. The Cryosphere, 17 (2). pp. 499-518. ISSN 1994-0424 cc_by_4_0 CC-BY F800 Physical and Terrestrial Geographical and Environmental Sciences Article PeerReviewed 2023 ftunivnorthumb https://doi.org/10.5194/tc-17-499-2023 2023-02-09T23:31:16Z Through their role in buttressing upstream ice flow, Antarctic ice shelves play an important part in regulating future sea-level change. Reduction in ice-shelf buttressing caused by increased ocean-induced melt along their undersides is now understood to be one of the key drivers of ice loss from the Antarctic ice sheet. However, despite the importance of this forcing mechanism, most ice-sheet simulations currently rely on simple melt parameterisations of this ocean-driven process since a fully coupled ice–ocean modelling framework is prohibitively computationally expensive. Here, we provide an alternative approach that is able to capture the greatly improved physical description of this process provided by large-scale ocean-circulation models over currently employed melt parameterisations but with trivial computational expense. This new method brings together deep learning and physical modelling to develop a deep neural network framework, MELTNET, that can emulate ocean model predictions of sub-ice-shelf melt rates. We train MELTNET on synthetic geometries, using the NEMO ocean model as a ground truth in lieu of observations to provide melt rates both for training and for evaluation of the performance of the trained network. We show that MELTNET can accurately predict melt rates for a wide range of complex synthetic geometries, with a normalised root mean squared error of 0.11 m yr−1 compared to the ocean model. MELTNET calculates melt rates several orders of magnitude faster than the ocean model and outperforms more traditional parameterisations for > 96 % of geometries tested. Furthermore, we find MELTNET's melt rate estimates show sensitivity to established physical relationships such as changes in thermal forcing and ice-shelf slope. This study demonstrates the potential for a deep learning framework to calculate melt rates with almost no computational expense, which could in the future be used in conjunction with an ice sheet model to provide predictions for large-scale ice sheet models. Article in Journal/Newspaper Antarc* Antarctic Ice Sheet Ice Shelf Ice Shelves The Cryosphere Northumbria University, Newcastle: Northumbria Research Link (NRL) Antarctic The Antarctic The Cryosphere 17 2 499 518 |
institution |
Open Polar |
collection |
Northumbria University, Newcastle: Northumbria Research Link (NRL) |
op_collection_id |
ftunivnorthumb |
language |
English |
topic |
F800 Physical and Terrestrial Geographical and Environmental Sciences |
spellingShingle |
F800 Physical and Terrestrial Geographical and Environmental Sciences Rosier, Sebastian Bull, Christopher Woo, Wai Lok Gudmundsson, Hilmar Predicting ocean-induced ice-shelf melt rates using deep learning |
topic_facet |
F800 Physical and Terrestrial Geographical and Environmental Sciences |
description |
Through their role in buttressing upstream ice flow, Antarctic ice shelves play an important part in regulating future sea-level change. Reduction in ice-shelf buttressing caused by increased ocean-induced melt along their undersides is now understood to be one of the key drivers of ice loss from the Antarctic ice sheet. However, despite the importance of this forcing mechanism, most ice-sheet simulations currently rely on simple melt parameterisations of this ocean-driven process since a fully coupled ice–ocean modelling framework is prohibitively computationally expensive. Here, we provide an alternative approach that is able to capture the greatly improved physical description of this process provided by large-scale ocean-circulation models over currently employed melt parameterisations but with trivial computational expense. This new method brings together deep learning and physical modelling to develop a deep neural network framework, MELTNET, that can emulate ocean model predictions of sub-ice-shelf melt rates. We train MELTNET on synthetic geometries, using the NEMO ocean model as a ground truth in lieu of observations to provide melt rates both for training and for evaluation of the performance of the trained network. We show that MELTNET can accurately predict melt rates for a wide range of complex synthetic geometries, with a normalised root mean squared error of 0.11 m yr−1 compared to the ocean model. MELTNET calculates melt rates several orders of magnitude faster than the ocean model and outperforms more traditional parameterisations for > 96 % of geometries tested. Furthermore, we find MELTNET's melt rate estimates show sensitivity to established physical relationships such as changes in thermal forcing and ice-shelf slope. This study demonstrates the potential for a deep learning framework to calculate melt rates with almost no computational expense, which could in the future be used in conjunction with an ice sheet model to provide predictions for large-scale ice sheet models. |
format |
Article in Journal/Newspaper |
author |
Rosier, Sebastian Bull, Christopher Woo, Wai Lok Gudmundsson, Hilmar |
author_facet |
Rosier, Sebastian Bull, Christopher Woo, Wai Lok Gudmundsson, Hilmar |
author_sort |
Rosier, Sebastian |
title |
Predicting ocean-induced ice-shelf melt rates using deep learning |
title_short |
Predicting ocean-induced ice-shelf melt rates using deep learning |
title_full |
Predicting ocean-induced ice-shelf melt rates using deep learning |
title_fullStr |
Predicting ocean-induced ice-shelf melt rates using deep learning |
title_full_unstemmed |
Predicting ocean-induced ice-shelf melt rates using deep learning |
title_sort |
predicting ocean-induced ice-shelf melt rates using deep learning |
publisher |
Coperincus |
publishDate |
2023 |
url |
https://nrl.northumbria.ac.uk/id/eprint/51332/ https://doi.org/10.5194/tc-17-499-2023 https://nrl.northumbria.ac.uk/id/eprint/51332/1/tc_17_499_2023.pdf |
geographic |
Antarctic The Antarctic |
geographic_facet |
Antarctic The Antarctic |
genre |
Antarc* Antarctic Ice Sheet Ice Shelf Ice Shelves The Cryosphere |
genre_facet |
Antarc* Antarctic Ice Sheet Ice Shelf Ice Shelves The Cryosphere |
op_relation |
https://nrl.northumbria.ac.uk/id/eprint/51332/1/tc_17_499_2023.pdf Rosier, Sebastian, Bull, Christopher, Woo, Wai Lok and Gudmundsson, Hilmar (2023) Predicting ocean-induced ice-shelf melt rates using deep learning. The Cryosphere, 17 (2). pp. 499-518. ISSN 1994-0424 |
op_rights |
cc_by_4_0 |
op_rightsnorm |
CC-BY |
op_doi |
https://doi.org/10.5194/tc-17-499-2023 |
container_title |
The Cryosphere |
container_volume |
17 |
container_issue |
2 |
container_start_page |
499 |
op_container_end_page |
518 |
_version_ |
1766254629874565120 |