Molards as proxies of mountain permafrost degradation: direct comparison of experimental studies and field observations

Mountain permafrost is increasingly retreating due to climate change. This retreat leads to positive climatic feedback loops and poses safety risks due to more frequent slope instabilities. Therefore, assessing the condition and evolution of permafrost is critical. However, mapping the extent and retreat of permafrost is not as straightforward as for other elements of the cryosphere because permafrost cannot be directly mapped by remote sensing. 
In some mountain landslides there are cones of loose debris, which are remnants of formerly ice-cemented blocks. These cones are called “molards” and they indicate the presence of an area of discontinuous permafrost at the level of the detachment zone. The initial ice-cemented blocks range in height from 50 cm to 15 meters. 

The goal of this project is to use molards as proxies of mountain permafrost degradation. Therefore, we have to understand the physical processes leading to the formation of molards as well as how these processes determine the final molard shape. 
To achieve this goal we recreate molards by using physical modeling and we have investigated molards at several Icelandic field sites. For the physical modeling it is necessary to downscale the molards to an initial cube size of ~30 cm due to current laboratory limitations. The initial blocks are created by freezing fully water saturated sediment in a wooden mold at -20°C for 48 hours. 
Sediment from actual Icelandic molards is used as well as other reduced complexity simulants with different grain sizes, grain shapes, and clay content.
We let the blocks degrade for 72 hours under a controlled and monitored laboratory environment with constant temperature and humidity conditions. We use a photogrammetric time-lapse system to create a digital elevation model of the degrading block to detect changes in hourly time-steps. 

Our initial results show that increasing clay content strongly influences the degradation speed and the final molard shape because it increases cohesion. In the field we have identified conical and trapezoidal cross-sections as the predominant shape for molards. But in the laboratory setting, high clay content means that the blocks do not degrade into this characteristic shape (without further meteorological influence). In this case,  landslide-like processes and single rockfall events dominate the molard formation process. 
For coarser grain sizes and low clay contents, rockfall is the dominant process, and both the conical and trapezoidal cross-sections can be reproduced in the experiments.