摘要：To develop a better predictive tool for dispersion in rivers over a range of temporal and spatial scales, our group has developed a simple Lagrangian model that is applicable for a wide range of coordinate systems and flow modeling methodologies. The approach allows dispersion computations for a large suite of discretizations, model dimensions (1-, 2-, or 3-dimensional), spatial and temporal discretization, and turbulence closures. As the model is based on a discrete non-interacting particle approach, parallelization is straightforward, such that simulations with large numbers of particles are tractable. Results from the approach are compared to dispersion measurements made with conventional Rhodamine WT dye experiment in which typical at-a-point sensors are employed to determine concentration. The model performs well, but spatial resolution for experiments over large and or complex river flows was inadequate for model testing. To address this issue, we explored the idea of measuring spatial concentrations in river flows using hyperspectral remote sensing. Experiments both for idealized channels and real rivers show that this technique is viable and can provide high levels of spatial detail in concentration measurements with quantitatively accurate concentrations.
其他摘要：To develop a better predictive tool for dispersion in rivers over a range of temporal and spatial scales, our group has developed a simple Lagrangian model that is applicable for a wide range of coordinate systems and flow modeling methodologies. The approach allows dispersion computations for a large suite of discretizations, model dimensions (1-, 2-, or 3-dimensional), spatial and temporal discretization, and turbulence closures. As the model is based on a discrete non-interacting particle approach, parallelization is straightforward, such that simulations with large numbers of particles are tractable. Results from the approach are compared to dispersion measurements made with conventional Rhodamine WT dye experiment in which typical at-a-point sensors are employed to determine concentration. The model performs well, but spatial resolution for experiments over large and or complex river flows was inadequate for model testing. To address this issue, we explored the idea of measuring spatial concentrations in river flows using hyperspectral remote sensing. Experiments both for idealized channels and real rivers show that this technique is viable and can provide high levels of spatial detail in concentration measurements with quantitatively accurate concentrations.
