Determining Local Mesh Resolution for Accurate Modeling of River-Floodplain Connectivity

Abstract:

The river-floodplain system plays a fundamental role in fluvial ecosystem preservation, landscape evolution, and building resilience against natural and anthropogenic modifications. Remotely sensed data, such as high-resolution topography acquired with lidar, typically reveal remarkably complex systems of floodplain channels. These channels have been shown to convey water and facilitate river-floodplain connectivity at discharges lower than bankfull. In this study, we focus on the quantification of river-floodplain connectivity using a backwater-influenced reach of the lower Trinity River in Texas. Our approach combines field observations with the ANUGA hydrodynamic model to assess surface water connectivity and residence time in floodplain channels during Tropical Storm Imelda (2019). Numerical modeling of connectivity is often limited by mesh resolution, in the common case where floodplain features are smaller than the local element size. Channels can have widths on the order of 1-10 m, which is smaller than the mesh resolution often used to model a domain of this size in a computationally-efficient way. We utilize the flexibility of the unstructured ANUGA mesh to increase mesh resolution locally over the extent of three bends on the Trinity River and determine the most effective mesh scales for capturing the observed connectivity between the river and its floodplain. Mesh resolution along the levees is increased from a uniform resolution equal to about one-fifth of the river width to scales smaller than typical levee channels and similar in scale to the underlying DEM (1 m). We compare modeling results to field data collected from the Trinity River floodplain during Imelda, including stage, flow velocity, and flow direction. We then employ Lagrangian particle routing to quantify water residence time using various particle injection points and times. Results inform on the appropriate scales for quantifying hydrological connectivity as observed in the field, thus providing a tool for quantifying lateral exchange, residence time, pluvial flooding, and other factors related to river-floodplain connectivity.