Rock glaciers (RG) are landforms that occur in high latitudes or elevations and — in their active state — consist of a mixture of rock debris and ice. Despite serving as a form of groundwater storage, they are an indicator of the occurrence of (former) permafrost and therefore carry significance in the research for ongoing climate change.
For these reasons, the past years have shown rising interest in the establishment of RG inventories to investigate the extent of permafrost and quantify water storage. Creating these inventories, however, usually involves manual, laborious, and subjective mapping of the landforms based on aerial images - and digital elevation model analysis.
A novel way of creating RG inventories is to replace the human mapper with a Machine Learning algorithm that is trained to automatically detect RGs. In a previous study by Erharter et al. (2022), deep artificial neural networks (ANN) were successfully trained to map RGs in the Austrian Alps.
The goal of this master's thesis is therefore to apply the same workflow as in Erharter et al. (2022) and create a first RG inventory that covers all of Norway. Establishing such a RG inventory would be a milestone for Norwegian Quaternary geological and hydrogeological research as well as paving the way for several follow-up studies in these fields including revisions of existing small inventories (see image below).
Click here for a bigger version.
Erharter, G.H., Wagner, T., Winkler, G., Marcher, T., 2022. Machine learning – An approach for consistent rock glacier mapping and inventorying – Example of Austria. Applied Computing and Geosciences 16, 100093.