Are all runoff processes the same? Numerical experiments comparing a Darcy‐Richards solver to an overland flow‐based approach for subsurface storm runoff simulation

Abstract

AbstractHillslope runoff theory is based largely on the differentiation between infiltration excess overland flow, saturation excess overland flow, and subsurface stormflow. Here we explore to what extent a 2‐D friction‐based overland flow model is useful for predicting hillslope‐scale subsurface stormflow, posited here as phenomenologically the same as infiltration excess at depth. We compare our results to a 3‐D variably saturated Darcy‐Richards subsurface solver for individual rainfall runoff events. We use field data from the well‐studied Panola Mountain Experimental hillslope in Georgia USA. Our results show that the two models are largely indistinguishable in terms of their ability to simulate the hillslope hydrograph magnitude and timing for a range of slopes and rainfall depths. Furthermore, we find that the descriptive ability of the overland flow model is comparable to the variably saturated subsurface flow model in terms of its ability to represent the spatial distribution of subsurface stormflow and infiltration across the soil‐bedrock interface. More importantly, these results imply that the physics of infiltration excess subsurface stormflow at the soil‐bedrock interface is similar to infiltration excess overland flow at the soil surface, in terms of detention storage, loss along the lower boundary, and threshold‐like activation at the larger hillslope scale. Given the phenomenological similarity of overland flow and subsurface stormflow and the fact that overland flow model predictions are considerably faster to run (particularly as slope and rainfall depth increase), these findings imply that new forms of hillslope‐scale subsurface storm runoff predictions may be possible with the knowledge of bedrock permeability and limited soil information. Finally, this work suggests that the role of soil mantle vis‐à‐vis subsurface stormflow is mainly as a filter that delays the development of patches of saturation along the bedrock surface. Our model results show that simple realizations of soil based on a few soil depth measurements can possibly be enough to characterize this filtering effect.