Transport and retention of a conservative tracer in an isolated creek-marsh system

A study of tracer transport and retention in a small tidal creek and marsh system located in the southeastern US was conducted using a three-dimensional hydrodynamic model and data from a dye tracer release. The model simulated tidally driven flow, including inundation and drying out of the marshes and the dispersal of the dye tracer. Flow measurements in the tidal creek showed that the simulations appeared to generally duplicate the tidally driven flow into and out of the tidal creeks and marshes. The dye tracer experiment was conducted to test the hydrodynamic model's ability to simulate dispersal of a point release of pollutants into the creek during an incoming tide. The simulations reproduced much of the experimental measurements, but bathymetric errors and lack of resolution of the smallest arms of the tidal creeks affected the ability to faithfully reproduce the initial peak in measured dye tracer concentrations at a sampling location far from the boundary. The lack of the smallest tidal creeks led to some trapping of water and dye in the marshes. Two independent estimates of flushing time yielded values between 1.6 and 2.4 days. An uncertainty analysis indicates that model simulations are sensitive to variations in parameters such as water depth and marsh grass density. On the other hand, omission of the smallest tidal creeks in simulations may be partially offset by decreasing marsh grass density. Further improvements must rely on more accurate and detailed bathymetric data that resolves the smaller arms of the tidal creeks and includes quantitative information about the density and distribution of marsh grass. (C) 2010 Elsevier Ltd. All rights reserved.