Where the White Continent Is Blue: Deep Learning Locates Bare Ice in Antarctica
Abstract In some areas of Antarctica, blue‐colored bare ice is exposed at the surface. These blue ice areas (BIAs) can trap meteorites or old ice and are vital for understanding the climatic history. By combining multi‐sensor remote sensing data (MODIS, RADARSAT‐2, and TanDEM‐X) in a deep learning f...
Main Authors: | , , , , , , , |
---|---|
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
2024
|
Subjects: | |
Online Access: | http://hdl.handle.net/2013/ULB-DIPOT:oai:dipot.ulb.ac.be:2013/368812 https://dipot.ulb.ac.be/dspace/bitstream/2013/368812/3/Tollenaar2024_GRL.pdf |
id |
ftunivbruxelles:oai:dipot.ulb.ac.be:2013/368812 |
---|---|
record_format |
openpolar |
spelling |
ftunivbruxelles:oai:dipot.ulb.ac.be:2013/368812 2024-04-28T08:00:22+00:00 Where the White Continent Is Blue: Deep Learning Locates Bare Ice in Antarctica Tollenaar, Veronica Zekollari, Harry Pattyn, Frank Rußwurm, Marc Kellenberger, Benjamin Lhermitte, Stef Izeboud, Maaike Tuia, Devis 2024-02 1 full-text file(s): application/pdf http://hdl.handle.net/2013/ULB-DIPOT:oai:dipot.ulb.ac.be:2013/368812 https://dipot.ulb.ac.be/dspace/bitstream/2013/368812/3/Tollenaar2024_GRL.pdf en eng uri/info:doi/10.1029/2023GL106285 uri/info:scp/85184165035 https://dipot.ulb.ac.be/dspace/bitstream/2013/368812/3/Tollenaar2024_GRL.pdf http://hdl.handle.net/2013/ULB-DIPOT:oai:dipot.ulb.ac.be:2013/368812 1 full-text file(s): info:eu-repo/semantics/openAccess Geophysical research letters, 51 (3 Sciences exactes et naturelles Antarctica blue ice deep learning noisy labels info:eu-repo/semantics/article info:ulb-repo/semantics/articlePeerReview info:ulb-repo/semantics/openurl/article 2024 ftunivbruxelles 2024-04-10T00:08:55Z Abstract In some areas of Antarctica, blue‐colored bare ice is exposed at the surface. These blue ice areas (BIAs) can trap meteorites or old ice and are vital for understanding the climatic history. By combining multi‐sensor remote sensing data (MODIS, RADARSAT‐2, and TanDEM‐X) in a deep learning framework, we map blue ice across the continent at 200‐m resolution. We use a novel methodology for image segmentation with “noisy” labels to learn an underlying “clean” pattern with a neural network. In total, BIAs cover ca. 140,000 km 2 (∼1%) of Antarctica, of which nearly 50% located within 20 km of the grounding line. There, the low albedo of blue ice enhances melt‐water production and its mapping is crucial for mass balance studies that determine the stability of the ice sheet. Moreover, the map provides input for fieldwork missions and can act as constraint for other geophysical mapping efforts. SCOPUS: ar.j info:eu-repo/semantics/published Article in Journal/Newspaper Antarc* Antarctica Ice Sheet DI-fusion : dépôt institutionnel de l'Université libre de Bruxelles (ULB) |
institution |
Open Polar |
collection |
DI-fusion : dépôt institutionnel de l'Université libre de Bruxelles (ULB) |
op_collection_id |
ftunivbruxelles |
language |
English |
topic |
Sciences exactes et naturelles Antarctica blue ice deep learning noisy labels |
spellingShingle |
Sciences exactes et naturelles Antarctica blue ice deep learning noisy labels Tollenaar, Veronica Zekollari, Harry Pattyn, Frank Rußwurm, Marc Kellenberger, Benjamin Lhermitte, Stef Izeboud, Maaike Tuia, Devis Where the White Continent Is Blue: Deep Learning Locates Bare Ice in Antarctica |
topic_facet |
Sciences exactes et naturelles Antarctica blue ice deep learning noisy labels |
description |
Abstract In some areas of Antarctica, blue‐colored bare ice is exposed at the surface. These blue ice areas (BIAs) can trap meteorites or old ice and are vital for understanding the climatic history. By combining multi‐sensor remote sensing data (MODIS, RADARSAT‐2, and TanDEM‐X) in a deep learning framework, we map blue ice across the continent at 200‐m resolution. We use a novel methodology for image segmentation with “noisy” labels to learn an underlying “clean” pattern with a neural network. In total, BIAs cover ca. 140,000 km 2 (∼1%) of Antarctica, of which nearly 50% located within 20 km of the grounding line. There, the low albedo of blue ice enhances melt‐water production and its mapping is crucial for mass balance studies that determine the stability of the ice sheet. Moreover, the map provides input for fieldwork missions and can act as constraint for other geophysical mapping efforts. SCOPUS: ar.j info:eu-repo/semantics/published |
format |
Article in Journal/Newspaper |
author |
Tollenaar, Veronica Zekollari, Harry Pattyn, Frank Rußwurm, Marc Kellenberger, Benjamin Lhermitte, Stef Izeboud, Maaike Tuia, Devis |
author_facet |
Tollenaar, Veronica Zekollari, Harry Pattyn, Frank Rußwurm, Marc Kellenberger, Benjamin Lhermitte, Stef Izeboud, Maaike Tuia, Devis |
author_sort |
Tollenaar, Veronica |
title |
Where the White Continent Is Blue: Deep Learning Locates Bare Ice in Antarctica |
title_short |
Where the White Continent Is Blue: Deep Learning Locates Bare Ice in Antarctica |
title_full |
Where the White Continent Is Blue: Deep Learning Locates Bare Ice in Antarctica |
title_fullStr |
Where the White Continent Is Blue: Deep Learning Locates Bare Ice in Antarctica |
title_full_unstemmed |
Where the White Continent Is Blue: Deep Learning Locates Bare Ice in Antarctica |
title_sort |
where the white continent is blue: deep learning locates bare ice in antarctica |
publishDate |
2024 |
url |
http://hdl.handle.net/2013/ULB-DIPOT:oai:dipot.ulb.ac.be:2013/368812 https://dipot.ulb.ac.be/dspace/bitstream/2013/368812/3/Tollenaar2024_GRL.pdf |
genre |
Antarc* Antarctica Ice Sheet |
genre_facet |
Antarc* Antarctica Ice Sheet |
op_source |
Geophysical research letters, 51 (3 |
op_relation |
uri/info:doi/10.1029/2023GL106285 uri/info:scp/85184165035 https://dipot.ulb.ac.be/dspace/bitstream/2013/368812/3/Tollenaar2024_GRL.pdf http://hdl.handle.net/2013/ULB-DIPOT:oai:dipot.ulb.ac.be:2013/368812 |
op_rights |
1 full-text file(s): info:eu-repo/semantics/openAccess |
_version_ |
1797572629258829824 |