Modelling soil creep dynamics based on meteoric and terrestrial Be-10 coupling: validation on a hillslope profile in northern Abukuma Mountain, Japan

                 Soils on hillslopes in humid temperate regions move slowly downward with varying fluxes depending on depth. The soil creep contributes to enhance soil depth in head hollows and develop soil layer structures in a long time scale, which regulate short-term subsurface hydrological processes and hence slope instability by rainwater infiltration. This study attempted to model long-term soil creep processes, based both on terrestrial Be-10 concentrations in bedrock and depth profiles of meteoric Be-10 abundance in soil columns on a hillslope. The soil creep model developed here composes of two velocity profiles corresponding to viscous transport by plastic deformation of soil mass and discrete particle-flow transport due to the external disturbance from the soil surface. By coupling these distinct transport laws, the depth profile of soil creep rates was modeled as a combined formula of two exponential functions. Soil production function determined by terrestrial Be-10 and spatial distribution of soil depths combined with topographic curvature provided the rate and mass-budget constraints for the model, while meteoric Be-10 inventory in the soil columns served a clue to fix model parameters for the subsurface soil creep pattern. The model fitting to the meteoric Be-10 profiles output the soil residence time, which was then examined by C-14 dating for the charcoals obtained from the soil column.

             Model validation was conducted at a hillslope underlain by granodiorite in northern Abukuma Mountains, Japan. The target hillslope exhibits a convex ridge top with a planar midsection that becomes slightly concave toward downslope. A thick soil layer (>1 m) covers the entire hillslope, which characteristically composed of two layers: upper soft and lower stiff parts divided around 40–60 cm in depth. Bedrocks were sampled at bottoms of several soil pits excavated on hill-noses for terrestrial Be-10 measurement. Soil samples for meteoric Be-10 analysis were collected sequentially from ground surface to ~2 m depth in 10 cm interval at four pits located along a transect from the ridge to hollow. Charcoal grains imbedded in the soil layer were also collected for C-14 dating. The datasets of meteoric Be-10 profiles were well explained by fitting the two-layered soil creep model under the terrestrial Be-10 derived soil production rates and depth-dependent soil residence time evidenced by the charcoal ages.