Automated extraction of glacier calving front locations from optical satellite imagery using deep learning

Iceberg calving has a strong impact on the stresses of outlet glaciers and their discharge. Representing these glacier dynamics is an essential part of constraining the glacial evolution and considerably impacts simulation results when projecting future sea-level contributions. The increasing availability and quality of remote sensing imagery enable us to realize a continuous and precise mapping of relevant parameters such as calving front locations. However, the huge amount of data accentuates the necessity for intelligent data analysis strategies.

This talk will present deep learning methodology for automatically detecting calving front margins in optical satellite imagery. The workflow is based on semantic image segmentation using a Convolutional Neural Network as well as a unique set of multi-spectral, textural and topographic input features. Jointly with the proposed toolkit, this talk presents an exceedingly dense dataset for 20 of the most important Greenlandic outlet glaciers.

  • When: Thursday, 15 April 2021 17:30
  • Who: MSc. Erik Loebel (Technische Universität Dresden, Professur für Geodätische Erdsystemforschung)
  • Where: Online BBB

Slides