Back to Search
Start Over
Calving Front Machine (CALFIN): glacial termini dataset and automated deep learning extraction method for Greenland, 1972–2019.
- Source :
- Cryosphere; Mar2021, Vol. 15 Issue 3, p1663-1675, 13p
- Publication Year :
- 2021
-
Abstract
- Sea level contributions from the Greenland Ice Sheet are influenced by the rapid changes in glacial terminus positions. The documentation of these evolving calving front positions, for which satellite imagery forms the basis, is therefore important. However, the manual delineation of these calving fronts is time consuming, which limits the availability of these data across a wide spatial and temporal range. Automated methods face challenges that include the handling of clouds, illumination differences, sea ice mélange, and Landsat 7 scan line corrector errors. To address these needs, we develop the Calving Front Machine (CALFIN), an automated method for extracting calving fronts from satellite images of marine-terminating glaciers, using neural networks. The results are often indistinguishable from manually curated fronts, deviating by on average 86.76 ± 1.43 m from the measured front. Landsat imagery from 1972 to 2019 is used to generate 22 678 calving front lines across 66 Greenlandic glaciers. This improves on the state of the art in terms of the spatiotemporal coverage and accuracy of its outputs and is validated through a comprehensive intercomparison with existing studies. The current implementation offers a new opportunity to explore subseasonal and regional trends on the extent of Greenland's margins and supplies new constraints for simulations of the evolution of the mass balance of the Greenland Ice Sheet and its contributions to future sea level rise. [ABSTRACT FROM AUTHOR]
- Subjects :
- DEEP learning
GREENLAND ice
SEA ice
ICE sheets
REMOTE-sensing images
Subjects
Details
- Language :
- English
- ISSN :
- 19940416
- Volume :
- 15
- Issue :
- 3
- Database :
- Complementary Index
- Journal :
- Cryosphere
- Publication Type :
- Academic Journal
- Accession number :
- 149730810
- Full Text :
- https://doi.org/10.5194/tc-15-1663-2021