Modeling the influence of subsurface topography on spatial and temporal variability of subsurface stormflow

Recent investigations of spatial patterns of soil depth, water table and subsurface flow response at the hillslope scale suggest that the variability in depth to bedrock (or to any other low permeable layer) may be a primary control on the space-time variability of subsurface stormflow. However, simulating or even predicting subsurface flow variability is still a challenge. Even more problematic is the fact that spatially explicit soil depth information is generally lacking at hillslope and catchment scales. We investigate how soil depth variability affects the spatial and temporal response of subsurface stormflow and propose a new way forward to defining these controls on sites without soil depth data. We used long-term data of spatially explicit subsurface flow measurements from a trenched hillslope at the Panola Mountain Research Watershed, as well as soil depth and soil property information, to calibrate and verify the HillVi hillslope model. Then we generated numerous realizations of the subsurface topography using geo-statistical information of the observed soil depth in the watershed. Subsurface flow variability was simulated based on these subsurface topography realizations. The spatial-temporal properties of the modeled flow were compared with the spatial and temporal variability of the observed flow. HillVi is a quasi 3D spatially explicit saturated and unsaturated water balance model and is well suited for this approach as it captures all major subsurface runoff generation processes (matrix and macropore flow and infiltration, lateral subsurface flow and pipe flow). We discuss the potentials and drawbacks of this approach to simulate spatially variable outflow from hillslopes and the general role of subsurface topography on the spatial and temporal patterns of subsurface flow.