Preprints
https://doi.org/10.5194/egusphere-2024-223
https://doi.org/10.5194/egusphere-2024-223
11 Mar 2024
 | 11 Mar 2024
Status: this preprint is open for discussion.

Deep learning based automatic grounding line delineation in DInSAR interferograms

Sindhu Ramanath Tarekere, Lukas Krieger, Dana Floricioiu, and Konrad Heidler

Abstract. The regular and robust mapping of grounding lines is essential for various applications related to the mass balance of marine ice sheets and glaciers, especially in Antarctica and Greenland. With Differential Interferometric Synthetic Aperture Radar (DInSAR) interferograms, it is possible to accurately capture the tide-induced bending of the ice shelf at a continent-wide scale and a temporal resolution of a few days. While current processing chains typically automatically generate differential interferograms, grounding lines are still primarily identified and delineated on the interferograms by a human operator. This method is time-consuming and inefficient, considering the volume of data from current and future SAR missions. We developed a pipeline that utilizes the Holistically-Nested Edge Detection (HED) neural network to delineate DInSAR interferograms automatically. We trained HED in a supervised manner using 421 manually annotated grounding lines for outlet glaciers and ice shelves on the Antarctic Ice Sheet. We also assessed the contribution of non-interferometric features like elevation, ice velocity and differential tide levels towards the delineation task. Our best-performing network generated grounding lines with a median distance of 186 m from the manual delineations. Additionally, we applied the network to generate grounding lines for undelineated interferograms, demonstrating the network's generalization capabilities and potential to generate high-resolution temporal and spatial mappings.

Sindhu Ramanath Tarekere, Lukas Krieger, Dana Floricioiu, and Konrad Heidler

Status: open (until 17 May 2024)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
Sindhu Ramanath Tarekere, Lukas Krieger, Dana Floricioiu, and Konrad Heidler
Sindhu Ramanath Tarekere, Lukas Krieger, Dana Floricioiu, and Konrad Heidler

Viewed

Total article views: 197 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
141 46 10 197 6 8
  • HTML: 141
  • PDF: 46
  • XML: 10
  • Total: 197
  • BibTeX: 6
  • EndNote: 8
Views and downloads (calculated since 11 Mar 2024)
Cumulative views and downloads (calculated since 11 Mar 2024)

Viewed (geographical distribution)

Total article views: 212 (including HTML, PDF, and XML) Thereof 212 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 12 Apr 2024
Download
Short summary
Grounding lines are geophysical features that divide ice masses on the bedrock and floating ice shelves. Their accurate location is required for calculating the mass balance of ice sheets and glaciers in Antarctica and Greenland. Human experts still manually detect them in satellite-based interferometric radar images, which is inefficient given the growing volume of data. We have developed an artificial intelligence-based automatic detection algorithm to generate Antarctic-wide grounding lines.