The influence of the Critical Zone on hillslope hydrology and stability

The Earth's near-surface layer, which extends from the vegetation canopy to unweathered bedrock, is known as the Critical Zone (CZ). It is essential for regulating the incidence of hydrological processes and hydro-geo hazards. Enhancing predictive models of infiltration, discharge production, and landslide initiation, requires an understanding of the relationships between the CZ and hillslope hydrology. In this research, the impact of changes in CZ configuration and characteristics on hillslope hydrology and stability is examined in various geomorphological and climatic contexts. The research begins by investigating rainfall infiltration mechanisms in the stratified pyroclastic soil-mantled slopes of the Campania region, Southern Italy. Field monitoring, laboratory infiltration experiments, and physically based hydrological modeling were combined to reproduce observed pore-water pressures, soil-moisture dynamics, and subsurface flow mechanisms. The findings enhanced our understanding of rainfall-induced slope instability and provided further insights into the redistribution of water in layered volcanic deposits. The work expands by using a coupled hydro-mechanical framework to forecast the onset of shallow landslides. The framework combines i) a transient hydrological model with ii) a finite element solver for slope stability in order to iii) compute local field factors of safety (LFS) under dynamic rainfall conditions. The framework's ability to faithfully replicate observed hillslope failures is validated against a benchmark test and actual storm events in the Braies Alpine Catchment, Alto Adige, Northern Italy. This supports the framework's use as a large- scale, cost-effective tool for quantitative hazard assessment in mountainous regions. The focus then shifts to exploring how the internal structure of the CZ controls hydrological response and slope stability, using the Coos Bay experimental catchment CB1 in Oregon, USA, as a test case. Extensive field data supported hydrological simulations of four distinct CZ configurations, ranging from soil-only to scenarios including variably weathered bedrock layers shaped by topographic and stress-field controls. Simulations of both controlled sprinklers and natural storm events revealed that the presence of a deep, irregularly weathered bedrock layer controls subsurface storage, groundwater movements, and localized pressure build-up, improving the agreement between simulated and observed streamflow and pore-pressure responses with respect to the commonly adopted soil-unweathered bedrock configuration. Lastly, the latter study was refined by adding a spatially variable field of hydraulic conductivity at CB1, which was constructed by seismic refraction data, hundreds of field slug tests, and predicted topographic stress fields. The value of combining geophysical constraints with hydromechanical modeling for a more realistic characterization of CZ behavior was demonstrated by the notable improvement in reproducing hydrological responses and slope failure processes derived from the improved characterization of CZ heterogeneity. All of these studies demonstrate that hillslope hydrology and stability are subject to first-order influences imposed by the CZ shape, stratification, and hydraulic characteristics. To provide more accurate prediction tools for hydrological forecasting and geohazard mitigation, this dissertation emphasizes the importance of explicitly representing CZ heterogeneity in physically-based hydrological analyses.