Fully integrated modeling of surface‐subsurface solute transport and the effect of dispersion in tracer hydrograph separation. Issue 10 (6th October 2014)

• Record Type: Journal Article
• Authors: Liggett, Jessica E.; Werner, Adrian D.; Smerdon, Brian D.; Partington, Daniel; Simmons, Craig T.
• Is Part Of: Water resources research. Volume 50:Issue 10(2014:Oct.)
• Page Start: 7750
• Page End: 7765
• Publication Date: 2014-10-06
• Subjects: Hydrology -- Periodicals; 333.91
• DOI: 10.1002/2013WR015040
• Languages: English
• ISSNs: 0043-1397

Abstract

<abstract abstract-type="main"> <title>Abstract</title> <p>Tracer hydrograph separation has been widely applied to identify streamflow components, often indicating that pre‐event water comprises a large proportion of stream water. Previous work using numerical modeling suggests that hydrodynamic mixing in the subsurface inflates the pre‐event water contribution to streamflow when derived from tracer‐based hydrograph separation. This study compares the effects of hydrodynamic dispersion, both within the subsurface and at the surface‐subsurface boundary, on the tracer‐based pre‐event water contribution to streamflow. Using a fully integrated surface‐subsurface code, we simulate two hypothetical 2‐D hillslopes with surface‐subsurface solute exchange represented by different solute transport conceptualizations (i.e., advective and dispersive conditions). Results show that when surface‐subsurface solute transport occurs via advection only, the pre‐event water contribution from the tracer‐based separation agrees well with the hydraulically determined value of pre‐event water from the numerical model, despite dispersion occurring within the subsurface. In this case, subsurface dispersion parameters have little impact on the tracer‐based separation results. However, the pre‐event water contribution from the tracer‐based separation is larger when dispersion at the surface‐subsurface boundary is considered. This work demonstrates that dispersion within the subsurface may not always be a significant factor in apparently large pre‐event water fluxes over a single rainfall event. Instead, dispersion at the surface‐subsurface boundary may increase estimates of pre‐event water contribution. This work also shows that solute transport in numerical models is highly sensitive to the representation of the surface‐subsurface interface. Hence, models of catchment‐scale solute dynamics require careful treatment and sensitivity testing of the surface‐subsurface interface to avoid misinterpretation of real‐world physical processes.</p> </abstract>