Wetlands offer protection for surface water quality by transforming and filtering a wide variety of water-borne contaminants. Yet, for many wetlands, we have a poor understanding of the physical transport that controls these functions. In particular, there are no models for dispersion in wetlands. This project combines an analytical framework with experimental observations to develop a model to predict dispersion in wetland systems. Once incorporated into transport models, the Broader Impact of this work will be to enable resource managers to predict changes in wetland function with land-use change, and to design constructed wetlands that mimic natural function. The Intellectual Merit will be the development and test of a new framework for predicting dispersion in vegetated flow, and in other multi-body or two-phase flows. Within a vegetated zone the velocity field is heterogeneous at the stem-scale. Directly behind each stem is a recirculation zone, in which the mean velocity is zero. Downstream of the recirculation zone is a wake in which the velocity is positive but diminished from the spatially averaged flow speed, U. Finally, between wakes and stems is a region of gap flow which, by conservation of mass, must be greater than U. Now consider a group of particles released together into the vegetated zone. Over time the particles become longitudinally dispersed, because each particle experiences a different series of velocities as it traverses multiple recirculation zones, wakes and gaps. The magnitude of the dispersion can be predicted from the following parameters that define the transport in each zone. 1) The mean residence time within the recirculation zones, 2) the size of the recirculation zones, 3) the stem drag coefficient, which defines the magnitude of the wake deficit and through continuity the gap augmentation, and 4) the lateral turbulent viscosity and diffusion. The first step in this study will be to measure these parameters in a laboratory model (cylinder array) over a range of flow and stem density found in the field. The second step will be to measure longitudinal dispersion in the cylinder array and compare the observed values to those predicted by theory. Application of the model to more complex canopies will be examined using dispersion measurements made in a marsh of Spartina alterniflora. Finally, we will combine the results from this study and EAR-0125056: Exchange Between Channel and Vegetated Zones to predict transport in partially vegetated channels and then test by observation.
