Author(s): Nguyen Quang Chien; Tan Soon Keat
Linked Author(s):
Keywords: Flow modelling; Quadtree grid; Everglades National Park
Abstract: Two-dimensional depth-averaged hydrodynamic models using unstructured grids for triangular meshes are widely used in the simulation of shallow water flows in coastal, estuarine area, and in floodplains hydrodynamics. However, automatic generation of triangular meshes is a complex process. In comparison, generation of quadtree grid is relatively less complex as it consists of the simple configuration of a Cartesian grid with a capability of local grid refinement to achieve higher resolution and detailed flow simulation. In this latter approach, the simulation domain is divided recursively, each square cell into four smaller quadrants, until the desired resolution (grid size) is reached. As a demonstration of the capability of the quadtree, this paper described the findings of a numerical experiment in which a quadtree grid was applied for 2-D flow modelling of Shark River Slough in Everglades National Park (ENP), Florida, USA. This is a mangrove ecosystem, where dense vegetation “ridges” (with high Manning roughness coefficient) obstruct the flow, causing gentle water surface slope with low flow velocity. The presence of tree islands (hammocks) in which the ground level is higher than water surface are treated as obstacles in the model. In this study, “Google TM Earth” was used as a pre-processor in which the tree island boundaries were drawn on, and then extracted to the calculation grid for hydrodynamic calculation. Water and earth surface elevation data are obtained from the EDEN[Everglades Depth Estimation Network]project. The simulation includes periods of high and low water level. The resulting water level and flow field provide the environmental flow field for evaluating the ecological conditions of the region.
Year: 2010