FathomDEM: an improved global terrain map using a hybrid vision transformer model
The Earth’s terrain is linked to many physical processes, and gaining the most accurate representation is key to work in many sectors from engineering to natural hazards modeling and ecology. Existing global digital elevation models (DEMs) are widely used, however often suffer from systematic biases...
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IOP Publishing
2025-01-01
|
| Series: | Environmental Research Letters |
| Subjects: | |
| Online Access: | https://doi.org/10.1088/1748-9326/ada972 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1823859156367441920 |
|---|---|
| author | Peter Uhe Chris Lucas Laurence Hawker Malcolm Brine Hamish Wilkinson Anthony Cooper Alex A Saoulis James Savage Christopher Sampson |
| author_facet | Peter Uhe Chris Lucas Laurence Hawker Malcolm Brine Hamish Wilkinson Anthony Cooper Alex A Saoulis James Savage Christopher Sampson |
| author_sort | Peter Uhe |
| collection | DOAJ |
| description | The Earth’s terrain is linked to many physical processes, and gaining the most accurate representation is key to work in many sectors from engineering to natural hazards modeling and ecology. Existing global digital elevation models (DEMs) are widely used, however often suffer from systematic biases caused by trees, buildings and instrumentation error, ultimately limiting their effectiveness. We present here, FathomDEM, a new global 30 m DEM produced using a novel application of a hybrid vision transformer model. This model removes surface artifacts from a global radar DEM, Copernicus DEM, aligning it more closely with true topography. In addition to improving on other global DEMs, FathomDEM also has reduced error compared to coastal-focussed DEMs such as the recent DeltaDTM. This demonstrates its impressive capacity to perform for specific landscapes, while being trained globally to model a wide range of terrain types. FathomDEM has been tested on the downstream task of flood modeling, showing increased accuracy compared to those run with the previous best global DEM, FABDEM, approaching the performance of LiDAR based flood modeling. This improvement is attributed to FathomDEM’s smaller error and substantial reduction in artifacts. This shows the suitability of FathomDEM for applied tasks and strengthens our evaluation compared to one based on vertical error alone. |
| format | Article |
| id | doaj-art-0d0ac386b7a14ba6a952e08f054ff513 |
| institution | Kabale University |
| issn | 1748-9326 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IOP Publishing |
| record_format | Article |
| series | Environmental Research Letters |
| spelling | doaj-art-0d0ac386b7a14ba6a952e08f054ff5132025-02-11T07:05:38ZengIOP PublishingEnvironmental Research Letters1748-93262025-01-0120303400210.1088/1748-9326/ada972FathomDEM: an improved global terrain map using a hybrid vision transformer modelPeter Uhe0https://orcid.org/0000-0003-4644-8559Chris Lucas1https://orcid.org/0009-0004-7972-3945Laurence Hawker2https://orcid.org/0000-0002-8317-7084Malcolm Brine3Hamish Wilkinson4https://orcid.org/0009-0001-3593-0502Anthony Cooper5https://orcid.org/0009-0009-6256-0429Alex A Saoulis6https://orcid.org/0009-0005-1486-8681James Savage7https://orcid.org/0000-0003-4807-0916Christopher Sampson8Fathom , Clifton Heights, Bristol, United KingdomFathom , Clifton Heights, Bristol, United KingdomFathom , Clifton Heights, Bristol, United Kingdom; School of Geographical Sciences, University of Bristol , Bristol, United Kingdom; Cabot Institute for the Environment, University of Bristol , Bristol, United KingdomFathom , Clifton Heights, Bristol, United KingdomFathom , Clifton Heights, Bristol, United KingdomFathom , Clifton Heights, Bristol, United KingdomFathom , Clifton Heights, Bristol, United Kingdom; Department of Physics & Astronomy, University College London , London, United KingdomFathom , Clifton Heights, Bristol, United KingdomFathom , Clifton Heights, Bristol, United KingdomThe Earth’s terrain is linked to many physical processes, and gaining the most accurate representation is key to work in many sectors from engineering to natural hazards modeling and ecology. Existing global digital elevation models (DEMs) are widely used, however often suffer from systematic biases caused by trees, buildings and instrumentation error, ultimately limiting their effectiveness. We present here, FathomDEM, a new global 30 m DEM produced using a novel application of a hybrid vision transformer model. This model removes surface artifacts from a global radar DEM, Copernicus DEM, aligning it more closely with true topography. In addition to improving on other global DEMs, FathomDEM also has reduced error compared to coastal-focussed DEMs such as the recent DeltaDTM. This demonstrates its impressive capacity to perform for specific landscapes, while being trained globally to model a wide range of terrain types. FathomDEM has been tested on the downstream task of flood modeling, showing increased accuracy compared to those run with the previous best global DEM, FABDEM, approaching the performance of LiDAR based flood modeling. This improvement is attributed to FathomDEM’s smaller error and substantial reduction in artifacts. This shows the suitability of FathomDEM for applied tasks and strengthens our evaluation compared to one based on vertical error alone.https://doi.org/10.1088/1748-9326/ada972terraindigital elevation modelmachine learningvision transformerremote sensing |
| spellingShingle | Peter Uhe Chris Lucas Laurence Hawker Malcolm Brine Hamish Wilkinson Anthony Cooper Alex A Saoulis James Savage Christopher Sampson FathomDEM: an improved global terrain map using a hybrid vision transformer model Environmental Research Letters terrain digital elevation model machine learning vision transformer remote sensing |
| title | FathomDEM: an improved global terrain map using a hybrid vision transformer model |
| title_full | FathomDEM: an improved global terrain map using a hybrid vision transformer model |
| title_fullStr | FathomDEM: an improved global terrain map using a hybrid vision transformer model |
| title_full_unstemmed | FathomDEM: an improved global terrain map using a hybrid vision transformer model |
| title_short | FathomDEM: an improved global terrain map using a hybrid vision transformer model |
| title_sort | fathomdem an improved global terrain map using a hybrid vision transformer model |
| topic | terrain digital elevation model machine learning vision transformer remote sensing |
| url | https://doi.org/10.1088/1748-9326/ada972 |
| work_keys_str_mv | AT peteruhe fathomdemanimprovedglobalterrainmapusingahybridvisiontransformermodel AT chrislucas fathomdemanimprovedglobalterrainmapusingahybridvisiontransformermodel AT laurencehawker fathomdemanimprovedglobalterrainmapusingahybridvisiontransformermodel AT malcolmbrine fathomdemanimprovedglobalterrainmapusingahybridvisiontransformermodel AT hamishwilkinson fathomdemanimprovedglobalterrainmapusingahybridvisiontransformermodel AT anthonycooper fathomdemanimprovedglobalterrainmapusingahybridvisiontransformermodel AT alexasaoulis fathomdemanimprovedglobalterrainmapusingahybridvisiontransformermodel AT jamessavage fathomdemanimprovedglobalterrainmapusingahybridvisiontransformermodel AT christophersampson fathomdemanimprovedglobalterrainmapusingahybridvisiontransformermodel |