Journal cover Journal topic
The Cryosphere An interactive open-access journal of the European Geosciences Union
Journal topic

Journal metrics

IF value: 4.713
IF4.713
IF 5-year value: 4.927
IF 5-year
4.927
CiteScore value: 8.0
CiteScore
8.0
SNIP value: 1.425
SNIP1.425
IPP value: 4.65
IPP4.65
SJR value: 2.353
SJR2.353
Scimago H <br class='widget-line-break'>index value: 71
Scimago H
index
71
h5-index value: 53
h5-index53
Preprints
https://doi.org/10.5194/tc-2020-74
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/tc-2020-74
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

  16 Apr 2020

16 Apr 2020

Review status
A revised version of this preprint was accepted for the journal TC.

DeepBedMap: Using a deep neural network to better resolve the bed topography of Antarctica

Wei Ji Leong and Huw Joseph Horgan Wei Ji Leong and Huw Joseph Horgan
  • Antarctic Research Centre, Victoria University of Wellington, Wellington, New Zealand

Abstract. To better resolve the bed elevation of Antarctica, we present DeepBedMap – a novel machine learning method that produces realistic Antarctic bed topography from multiple remote sensing data inputs. Our super-resolution deep convolutional neural network model is trained on scattered regions in Antarctica where high resolution (250 m) groundtruth bed elevation grids are available. The model is then used to generate high resolution bed topography in less well surveyed areas. DeepBedMap improves on previous interpolation methods by not restricting itself to a low spatial resolution (1000 m) BEDMAP2 raster image as its prior. It takes in additional high spatial resolution datasets, such as ice surface elevation, velocity and snow accumulation to better inform the bed topography even in the absence of ice-thickness data from direct ice-penetrating radar surveys. Our DeepBedMap model is based on an adapted Enhanced Super Resolution Generative Adversarial Network architecture, chosen to minimize per-pixel elevation errors while producing realistic topography. The final product is a four times upsampled (250 m) bed elevation model of Antarctica that can be used by glaciologists interested in the subglacial terrain, and by ice sheet modellers wanting to run catchment or continent-scale ice sheet model simulations. We show that DeepBedMap offers a more realistic topographic roughness profile compared to a standard bicubic interpolated BEDMAP2 and BedMachine Antarctica, and envision it to be used where a high resolution bed elevation model is required.

Wei Ji Leong and Huw Joseph Horgan

Interactive discussion

Status: final response (author comments only)
Status: final response (author comments only)
AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment
[Login for Authors/Editors] [Subscribe to comment alert] Printer-friendly Version - Printer-friendly version Supplement - Supplement

Wei Ji Leong and Huw Joseph Horgan

Data sets

DeepBedMap_DEM W. J. Leong and H. J. Horgan https://doi.org/10.17605/OSF.IO/96APW

Model code and software

DeepBedMap model W. J. Leong https://doi.org/10.5281/zenodo.3752614

Wei Ji Leong and Huw Joseph Horgan

Viewed

Total article views: 836 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
501 301 34 836 36 31
  • HTML: 501
  • PDF: 301
  • XML: 34
  • Total: 836
  • BibTeX: 36
  • EndNote: 31
Views and downloads (calculated since 16 Apr 2020)
Cumulative views and downloads (calculated since 16 Apr 2020)

Viewed (geographical distribution)

Total article views: 443 (including HTML, PDF, and XML) Thereof 440 with geography defined and 3 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 

Cited

Saved

No saved metrics found.

Discussed

No discussed metrics found.
Latest update: 23 Sep 2020
Publications Copernicus
Download
Short summary
A machine learning technique similar to the one used to enhance everyday photographs is applied to the problem of getting a better picture of Antarctica's bed – that which is hidden beneath the ice. By taking hints from what satellites can observe at the ice surface, the novel method learns to generate a realistic rough bed topography that complements existing approaches, with the result able to be used by scientists running fine scale ice sheet models relevant for predicting future sea level.
A machine learning technique similar to the one used to enhance everyday photographs is applied...
Citation