PROPOSAL NO.: CTS-0626164 PRINCIPAL INVESTIGATOR: EDWIN A. COWEN INSTITUTION: CORNELL UNIVERSITY
THE EFFECT OF SUBMERGED AND EMERGENT, HIGHLY FLEXIBLE AND RIGID MACROPHYTE CANOPY PATCHES ON FLOW AND MASS TRANSPORT
This grant provides funding to perform a series of carefully controlled experimental studies on the effects of aquatic macrophyte canopy patches on low-speed flows. The literature on the effects of terrestrial and aquatic vegetation on flow is becoming quite rich but there is a significant gap in the aquatic plant literature on the effects of heterogeneous canopies in low-speed flows and for the effects of highly flexible macrophtyes in general. The investigators will collect a canonical experimental data set documenting the effects of highly flexible macrophyte canopies and heterogenous macrophyte patches on low-speed flow and mass transport. A secondary goal is to extend and investigate the ability of existing turbulence closure schemes to capture these effects. The experiments will be conducted using two species of live macrophytes. The macrophytes will be placed in patches of increasing complexity in the all glass test section of a wide open-channel flume. State of the art quantitative imaging techniques will be used to measure the turbulent velocity characteristics, including the directly calculated turbulent dissipation, as well as the transport and dispersion of fluorescent tracers. Proposed modifications to turbulence closure models will be the starting point for investigating the ability of models to capture the turbulent kinetic energy budget. Competing numerical models of canopy flow that perform equally well in uniform aquatic canopies will be tested against this novel heterogeneous data set. The collected data set will fill critical gaps in our understanding of aquatic plant-flow interaction. It is important to work with live plants, as field biologists monitor macrophyte communities using metrics that do not lend themselves to scaling with simulated plant model results. The tested turbulence closure models will allow the development of computational tools capable of forecasting transport in natural environments dominated by both highly flexible and heterogeneous plant canopy patches - environments such as lakes of all scales, salt marshes, estuaries and coastal embayments and terrestrial flows as well. The project will lead to an interdisciplinary Ph.D. for a talented female student who will work with the PI to develop laboratory experiences for an introductory fluid mechanics course and to host weeklong projects for a program targeting high-school women interested in science and engineering.
Accurate prediction of pollutant transport and dilution in surface waters (such as lakes, rivers, and wetlands) is key to balancing the protection of freshwater resources with the economic interests of industry and municipalities. The complexity of surface water flows (often turbulent and forced by highly variable winds, air temperatures, and tributary inflows of differing densities) makes accurate prediction a challenge. In the second half of the twentieth century, as scientists and engineers advanced our basic understanding of turbulent flows, computer models were developed to predict velocities, mixing, and resulting pollutant transport and fate in surface waters. Today, computer models can account for the effects of inflows, winds, surface heating, complicated bed geometries, and temperature stratification on water velocities and turbulent mixing, but many frontiers remain. One of these frontiers is flow through aquatic vegetation, which is ubiquitous in coastal and inland waters (e.g., Figure 1a,b). The completed project produced a new computational model that predicts velocities, turbulent kinetic energy levels, and turbulent length scales, which are the most essential building blocks for predicting the spread of pollutants and nutrients in real aquatic vegetation. Previously existing models did not incorporate the length scale of plant stems, predicting identical turbulence levels for a 1mm pond weed and a 1m tree trunk. The new model incorporates the stem diameter. The new model was calibrated against a novel experimental data set, described in more detail below, from arrays of rigid cylinders having different diameters, and validated against an independent data set from submerged rigid cylinders and importantly, a new data set from live beds of Eurasian watermilfoil, an invasive species that is ubiquitous across the United States. As mentioned above, extensive laboratory data sets were collected on both rigid cylinder arrays, a simple model for rigid aquatic vegetation (Figure 2a), and live highly flexible plant canopies (Figure 2b). Given the difficulties with the estimation of drag induced by vegetation, a quantity important to model development, a new approach, known as a drag plate, was developed to directly measure the drag from the canopy, by instrumenting a platform mounted flush with the flume's bed, visible in Figure 2, allowing the measurement of the drag as a function of the parameters of the plant canopy (number of stems per area, flow speed, plant type, etc), including a unique use of quantitative imaging for the determination of time-varying frontal area. The drag plate proved to be an accurate and dependable experimental device. The research demonstrated that the strong vertical variation in plant frontal obstructed area leads to strong gradients in the velocity and turbulence fields as well as a strong vertical variation in mass dispersion processes. By using live plants, the group observed results qualitatively and quantitatively different to those found by other research groups, who have largely based their studies on plant surrogates, such as rigid dowels and dowels with flexible plastic model vegetation. The mean velocity and turbulence profiles indicate that the complex flows can be broken down into regions of more canonical flows such as shear flows and homogeneous turbulent flows that correlate to the plant morphology. Velocity results show a relatively higher speed region near the bed, compared to the no plants case, causing an increase in the bottom stress, which is important to sediment and larval dynamics. Strong gradients in the plant frontal area profile lead to a mixing layer around mid-depth, which has not been previously observed in emergent vegetation studies based on plant surrogates. The developed model of flow through aquatic vegetation, tested on both the new data sets collected as part of this project and previous data sets of others, is more soundly based on the fundamental physics of turbulent flow in natural vegetation than currently existing models, which allows it to better capture the broad range of features exhibited by these types of flows. The model naturally transitions between the submerged case, when plants are covered by a depth of water, to the emergent case, when the plants reach or protrude through the air-water interface; it is the only model designed to do this and it outperforms all other models, particularly at the prediction of the turbulent kinetic energy, a fundamental parameter that sets the rate of turbulent mixing. This new model will broadly extend the predictive capability of surface water models for pollutant transport to a wide range of vegetated coastal and inland surface water flows for which existing models do not make accurate predictions. Furthermore, the approach of separately modeling physical processes at their relevant scales to capture the multiple scales of turbulence generation is also appropriate for modeling flows through urban canopies, where the release of dangerous substances is a concern, or flow through wind farms, where the effects of the wakes of upstream wind turbines on down stream wind turbines is poorly understood.
