Influence of hydraulic conductivity variability on slope stability modelling in a first-order watershed

Abstract In landslide studies, the definition of geotechnical and hydrological parameters provides the basis for deterministic slope-stability modelling. However, their availability and spatial representativeness remain important limitations for watershed-scale applications. This study evaluated the influence of different saturated hydraulic conductivity (K sat ) spatialization approaches on Factor of Safety (FS) estimates using a deterministic slope-stability model in a first-order watershed. A Digital Terrain Model was used to derive morphometric and hydrological parameters, while field measurements of K sat supported three parameterization approaches: morphology-based, inverse distance weighting (IDW), and a single representative value. Instability scenarios were evaluated using a landslide scar inventory, and soil samples from six K sat measurement points were used to characterize particle-size distribution. Among the tested parameter configurations, maximum soil depth and initial groundwater-table depth produced greater differences in model results than K sat spatialization. K sat showed limited spatial variability, with measurements at 30 cm depth spanning from 10⁻³ to 10⁻⁵ m s⁻¹, whereas those at 50 cm were concentrated around 10⁻⁵ m s⁻¹. Higher clay contents with depth at some sampling points may have contributed to the observed decrease in K sat . Overall, within the observed range of K sat variability, a single representative value provided a practical alternative to spatially distributed parameterization in the studied first-order watershed. These findings are applicable to first-order watersheds where shallow landsliding is recurrent; broader applications involving greater geomorphological diversity and different predominant slope processes may require K sat characterization across a wider range of geomorphological compartments.