Geomorphic landslide susceptibility assessment using Slope-Area plots

Landslide susceptibility assessment is a key topic for land-use planning and for the overall safeguard of human activities. In this perspective, a wide range of methods and techniques have been proposed for the evaluation of landslide susceptibility, ranging from statistical methods to the latest deep learning technologies. Besides, Slope-Area plots are also exploited for the evaluation of surficial processes domains and the increasing availability of digital terrain models with higher resolution allows much detailed analyses.

On these premises, we compared slope-area plots produced with high resolution Lidar data with a landslide dataset produced following the LOOM data structure. The analysis has been carried out using only surficial phenomena like flows and falls. Moreover, such landslides have been decomposed into their principal components such as detachment, transit, and accumulation zones in order to perform an accurate evaluation of the geomorphic signature of such features. Each landslide has been also compared with the corresponding reference hillslope, defined as the set of enveloping Morse regions computed using Surface Network.

The plot of slope-area values of the training landslides and their reference hillslopes allows thresholding of the slope values at different contributing area bins, resulting in the mapping of those values exceeding the defined thresholds. Preliminary results show how such defined thresholds based on a proper training dataset could be a valid contribution to the overall topic of landslide susceptibility assessment based on geomorphological criteria at least for surficial landslide types like flow- and fall-like movements.