A field and modeling study of nonlinear storage‐discharge dynamics for an Alpine headwater catchment. Issue 2 (3rd February 2014)

• Record Type: Journal Article
• Authors: Camporese, Matteo; Penna, Daniele; Borga, Marco; Paniconi, Claudio
• Is Part Of: Water resources research. Volume 50:Issue 2(2014:Feb.)
• Page Start: 806
• Page End: 822
• Publication Date: 2014-02-03
• Subjects: Hydrology -- Periodicals; 333.91
• DOI: 10.1002/2013WR013604
• Languages: English
• ISSNs: 0043-1397

Abstract

<abstract abstract-type="main"> <title>Abstract</title> <p>A process‐based coupled model of surface‐subsurface flow is applied to the simulation of nonlinear hydrological dynamics for an experimental mountain headwater catchment in northeastern Italy. The comparison between measured and simulated responses, both distributed (water table and soil moisture) and integrated (streamflow at the outlet), shows that the model satisfactorily reproduces various nonlinear processes, in particular threshold behavior and hysteresis in the catchment storage‐discharge relationship. We typically observe a clockwise loop in this relationship, i.e., streamflow response is faster than groundwater and soil moisture response, due to larger time scales for subsurface processes and to soil moisture persistence and redistribution. The model is based on a standard Richards equation representation of integrated saturated‐unsaturated‐runoff dynamics and needs no ad hoc parameterization (e.g., for macropores, pipe flow, or retention curve hysteresis) to capture observed hysteretic relationships between storage and discharge. Additional numerical experiments are carried out to investigate how heterogeneity (bedrock permeability and the distinction between riparian and hillslope areas) and aquifer thickness and topography affect this nonlinear dynamics. The results show that catchment topography and soil depth exert the main control on the hysteresis and threshold patterns. This is evident from a spatial analysis of streamflow and water table response times to storm events, where the threshold points correspond to changes in terrain slope. These findings are confirmed by a further set of analyses carried out on an idealized v‐shaped catchment.</p> </abstract>