THE VALUE OF INCLUDING SOIL DEVELOPMENT PROCESSES IN MODELLING THE EVOLUTION OF LANDSLIDE-DOMINATED LANDSCAPES

The susceptibility for the occurrence of shallow landslides is among others a function of the properties of the soil or regolith. Beyond soil thickness, the angle of internal friction, wet bulk density and cohesion all co-determine the likelihood of sliding. Nonetheless, most existing landscape evolution models that include landslides lump these soil-related parameters together or set them as boundary conditions that are unaffected by the landslide themselves. This precludes simulation and analysis of the interplay between soil development on the one hand and landslide activity on the other hand.

To overcome this limitation, we have developed a landsliding module in soil-landscape evolution model LORICA that is based on existing formulation for the landsliding process, but connects them to dynamic rather than static soil properties. With this model, and based on a landscape and climate resembling a region in the Italian Apennines, we explored first the impact of landslide occurrence on the spatial patterns of soil over centennial timescales, second, the impact second the impact of soil development on on patterns of landslide occurrence, and third, whether the model’s representation of the mixed soil in the landslide deposit impacts time-space dynamics of landslides.

Results highlight the impact of landslides on transient but long-lived soil variability, the non-trivial impact of soil development on landslide patterns, and the different roles played by different soil properties in landslide path-dependency. Model code and executable are available as open-source materials, and we give a few suggestions about how to implement it in different settings.