@EGU General Assembly 2024

Water erosion is one of the most widespread forms of soil degradation and it is expected to increase as an effect of global change, especially due to the amplified frequency of extreme and localised rainfall events.
Focusing on the Mediterranean region, where these phenomena persist producing the most catastrophic erosion, this research aims to model soil erosion focusing on patterns and spatial distribution using rainfall data as detected by weather ground-radar upon an analysis of rainfall- cell patterns at regional scale (Tuscany, Italy).

Precipitations were elaborated by adopting a kriging method processing CAPPI (Constant Altitude Plan Position Indicator) reflectivity and rain gauges rainfall with an External Drift Kriging (KED) over a 30' time-step frames. Soil erosion was modelled using the platform based on the soil erosion/landscape evolution model LandSoil under Python coding simulating runoff combining a cinematic wave using St.Venant equations on a simplified “bucket” conceptual model for soil infiltration, and a two-level sediment redistribution process for rill and inter-rill erosion.

Soil erosion was simulated over a selection of rainfall events over the last ten years and results illustrated through a combined representation of factors that considers: 1. The sediment yield produced over the erosive event, 2. The land surface affected by the storm (i.e., rainfall-cell patterns of the event and subjected to runoff), 3. The involved rainfall energy and erosion rate of the processes. Such approach, applied to seasonal analysis, opens-up to a methodology better analysing spatial extent and effects on the ground. Drifts in erosion could be also detected by the use of long-terms rainfall series and return-time of the erosive events, useful to inspect evolution of the processes due to climate change.