Revisiting the Hewlett and Hibbert (1963) Hillslope Drainage Experiment and Modeling Effects of Decadal Pedogenic Processes and Leaky Soil Boundary Conditions

Abstract

AbstractSubsurface flow dominates water movement from hillslopes to streams in most forested headwater catchments. Hewlett and Hibbert (1963,https://doi.org/10.1029/JZ068i004p01081) constructed an idealized hillslope model (0.91 × 0.91 × 15.0 m; 21.8°) using reconstituted C horizon soil to investigate importance of interflow, a type of subsurface flow. They saturated the model, covered it to prevent evaporation, and allowed free drainage for 145 days. The resulting recession drainage curve suggested two phases: fast drainage of saturated soil in the first 1.5 days and then slow drainage of unsaturated soil. Hydrologists interpreted the latter as evidence interflow could sustain baseflow, even during extended drought. Since that experiment, typical forest vegetation grew in the model, providing root and litter inputs for 55 years. We removed all aboveground live biomass and repeated the experiment physically and numerically (HYDRUS‐2D), hypothesizing that pedogenesis would change the drainage curve and further elucidate the role of unsaturated flow from hillslopes. Contrary to this hypothesis, drainage curves in our twice‐repeated physical experiments and numerical simulation were unchanged for the first ~10 days, indicating pedogenesis and biological processes had not largely altered bulk hydraulic conductivities or soil moisture release characteristics. However, drainage unexpectedly ceased after about 2 weeks (14.3 ± 2.5 days), an order of magnitude sooner than in the original experiment, due to an apparent leak in the hillslope analogous to commonly observed bedrock fractures in natural systems. Thus, our results are a more natural recession behavior that highlight how incorporation of alternative hydrologic outputs can reduce drainage duration and volume from soils to baseflow.