Globally scalable glacier mapping by deep learning matches expert delineation accuracy
Abstract Accurate global glacier mapping is critical for understanding climate change impacts. Despite its importance, automated glacier mapping at a global scale remains largely unexplored. Here we address this gap and propose Glacier-VisionTransformer-U-Net (GlaViTU), a convolutional-transformer d...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-01-01
|
| Series: | Nature Communications |
| Online Access: | https://doi.org/10.1038/s41467-024-54956-x |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841559282380701696 |
|---|---|
| author | Konstantin A. Maslov Claudio Persello Thomas Schellenberger Alfred Stein |
| author_facet | Konstantin A. Maslov Claudio Persello Thomas Schellenberger Alfred Stein |
| author_sort | Konstantin A. Maslov |
| collection | DOAJ |
| description | Abstract Accurate global glacier mapping is critical for understanding climate change impacts. Despite its importance, automated glacier mapping at a global scale remains largely unexplored. Here we address this gap and propose Glacier-VisionTransformer-U-Net (GlaViTU), a convolutional-transformer deep learning model, and five strategies for multitemporal global-scale glacier mapping using open satellite imagery. Assessing the spatial, temporal and cross-sensor generalisation shows that our best strategy achieves intersection over union >0.85 on previously unobserved images in most cases, which drops to >0.75 for debris-rich areas such as High-Mountain Asia and increases to >0.90 for regions dominated by clean ice. A comparative validation against human expert uncertainties in terms of area and distance deviations underscores GlaViTU performance, approaching or matching expert-level delineation. Adding synthetic aperture radar data, namely, backscatter and interferometric coherence, increases the accuracy in all regions where available. The calibrated confidence for glacier extents is reported making the predictions more reliable and interpretable. We also release a benchmark dataset that covers 9% of glaciers worldwide. Our results support efforts towards automated multitemporal and global glacier mapping. |
| format | Article |
| id | doaj-art-5a80c3f6248c49bfbe5a7c984bf7d2d4 |
| institution | Kabale University |
| issn | 2041-1723 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Nature Communications |
| spelling | doaj-art-5a80c3f6248c49bfbe5a7c984bf7d2d42025-01-05T12:38:43ZengNature PortfolioNature Communications2041-17232025-01-0116111410.1038/s41467-024-54956-xGlobally scalable glacier mapping by deep learning matches expert delineation accuracyKonstantin A. Maslov0Claudio Persello1Thomas Schellenberger2Alfred Stein3Department of Earth Observation Science, Faculty of Geo-information Science and Earth Observation (ITC), University of TwenteDepartment of Earth Observation Science, Faculty of Geo-information Science and Earth Observation (ITC), University of TwenteDepartment of Geosciences, Faculty of Mathematics and Natural Sciences, University of OsloDepartment of Earth Observation Science, Faculty of Geo-information Science and Earth Observation (ITC), University of TwenteAbstract Accurate global glacier mapping is critical for understanding climate change impacts. Despite its importance, automated glacier mapping at a global scale remains largely unexplored. Here we address this gap and propose Glacier-VisionTransformer-U-Net (GlaViTU), a convolutional-transformer deep learning model, and five strategies for multitemporal global-scale glacier mapping using open satellite imagery. Assessing the spatial, temporal and cross-sensor generalisation shows that our best strategy achieves intersection over union >0.85 on previously unobserved images in most cases, which drops to >0.75 for debris-rich areas such as High-Mountain Asia and increases to >0.90 for regions dominated by clean ice. A comparative validation against human expert uncertainties in terms of area and distance deviations underscores GlaViTU performance, approaching or matching expert-level delineation. Adding synthetic aperture radar data, namely, backscatter and interferometric coherence, increases the accuracy in all regions where available. The calibrated confidence for glacier extents is reported making the predictions more reliable and interpretable. We also release a benchmark dataset that covers 9% of glaciers worldwide. Our results support efforts towards automated multitemporal and global glacier mapping.https://doi.org/10.1038/s41467-024-54956-x |
| spellingShingle | Konstantin A. Maslov Claudio Persello Thomas Schellenberger Alfred Stein Globally scalable glacier mapping by deep learning matches expert delineation accuracy Nature Communications |
| title | Globally scalable glacier mapping by deep learning matches expert delineation accuracy |
| title_full | Globally scalable glacier mapping by deep learning matches expert delineation accuracy |
| title_fullStr | Globally scalable glacier mapping by deep learning matches expert delineation accuracy |
| title_full_unstemmed | Globally scalable glacier mapping by deep learning matches expert delineation accuracy |
| title_short | Globally scalable glacier mapping by deep learning matches expert delineation accuracy |
| title_sort | globally scalable glacier mapping by deep learning matches expert delineation accuracy |
| url | https://doi.org/10.1038/s41467-024-54956-x |
| work_keys_str_mv | AT konstantinamaslov globallyscalableglaciermappingbydeeplearningmatchesexpertdelineationaccuracy AT claudiopersello globallyscalableglaciermappingbydeeplearningmatchesexpertdelineationaccuracy AT thomasschellenberger globallyscalableglaciermappingbydeeplearningmatchesexpertdelineationaccuracy AT alfredstein globallyscalableglaciermappingbydeeplearningmatchesexpertdelineationaccuracy |