Can a simple numerical model help to fine-tune the analysis of ground-penetrating radar data? Hochebenkar rock glacier as a case study

Little is known about the thickness of active Alpine rock glaciers, yet they are important components of the local hydrology. We use GPR data to determine the depth of the bedrock of Ausseres Hochebenkar rock glacier (Austria). There is no detailed information available regarding density and composition of the rock glacier, and assumptions about the signal propagation velocity have to be made when processing the GPR data. We use a simple creep model based on surface displacement and slope to calculate the thickness of the rock glacier along a flow line. We calculated bedrock profiles along the flow line for three different time periods, using input from multitemporal digital elevation models. We improved the fit of the profiles by calibrating the values used for layer densities and considered the model valid where the modelled bedrock profiles are within error of each other. We then compared the modeled values with the GPR data to check whether our assumptions for the propagation velocity produced results that match the model. While the fit is good at the lower end of the rock glacier, the GPR data appear to overestimate depth in the upper region. We adjusted the propagation velocity accordingly and find maximum thicknesses of over 50 m and a mean thickness of 30-40 m. The insights gained from the modeling approach thereby improved the fine-tuning of the GPR analysis.