VARIABILITY IN SOIL-WATER RETENTION PROPERTIES AND IMPLICATIONS FOR PHYSICS-BASED SIMULATION OF LANDSLIDE EARLY WARNING CRITERIA

Landsliding is a persistent hazard to human life and will likely remain so as populations continue to expand into failure-prone terrain. Despite the observed connection between the unsaturated zone and the shallow landslide initiation problem, there is considerable uncertainty in how the approaches used to estimate soil-water retention properties affect slope stability assessment. Further examination of this uncertainty is critical to evaluating the utility of hydrologic modeling as a physics-based tool for shallow landslide prediction and early warning. To examine conditions relevant to shallow landsliding, we employ a numerical model of variably saturated groundwater flow parameterized with an ensemble of texture-, laboratory-, and field-based soil-water retention properties for an extensively monitored landslide-prone site. Simulations of soil-water content (θ), pore-water pressure (u_w), and the resultant Factor of Safety (FS) show considerable variability across and within these different parameter estimation techniques. In particular, we demonstrate that variability in soil-water retention properties influences predictions of positive u_w coincident with widespread shallow landsliding. We also find that the ensemble of soil-water retention properties imposes an order-of-magnitude and two-fold variability in seasonal and event-scale landslide susceptibility, respectively. Despite reduced FS uncertainty during wet conditions, parameters that control the dry end of the soil-water retention function markedly impact the ability of a hydrologic model to capture θ dynamics observed in the field. These results suggest that neglecting variability in soil-water retention properties could be problematic for objective physics-based simulation of landslide early warning criteria.