Characterizing Compound Coastal-Riverine Behavior along the U.S. East Coast using a Coupled Hydrologic-Hydrodynamic Model

Roham Bakhtyar1, Kazungu Maitaria2, P. Velissariou3, Beheen Trimble4, Trey Flowers5, Saeed Moghimi6, Ali Abdolali7, Hassan Mashriqui8, Andre Jaco Van der Westhuysen9, Graeme R Aggett10 and Edward P Clark5, (1)NOAA / Office of Water Prediction / National Water Center, Tuscaloosa, AL, United States, (2)National Oceanic and Atmospheric Administration (NOAA), National Water Center (NWC), Tuscaloosa, AL, United States, (3)NOAA-NOS, Tuscaloosa, United States, (4)NOAA-NWC, Tuscaloosa, AL, United States, (5)NOAA Office of Water Prediction, National Water Center, Tuscaloosa, United States, (6)NOAA National Ocean Service, Silver Spring, MD, United States, (7)University Corporation for Atmospheric Research, College Park, MD, United States, (8)National Oceanic and Atmospheric Administration (NOAA), Silver Spring, United States, (9)NOAA Environmental Modeling Center, College Park, MD, United States, (10)Lynker, Boulder, CO, United States
Water dynamics along the U.S. East Coast are governed by multifaceted interactions of storm surge, tides, waves, freshwater inflows, winds and atmospheric pressure; hence, there is a need for an accurate coupled flood-modeling framework that suitably includes the physical processes in this area. The coupled model requires exchange of riverine information between the hydrodynamic and hydrologic models. In early studies, many of the flood inundation models ignored river discharge and, therefore, some important physical processes such as momentum exchange were lost. In some existing modeling frameworks, riverine discharge applied to hydrodynamic models are time-varying/constant hydrographs for the boundary condition. Recently, some Riverine-Estuarine models like the D-Flow FM model are capable of seamlessly integrating 2D physics with 1D hydraulics. In this study, the modeling system links the National Water Model with the D-Flow FM hydrodynamic model. Water levels and velocities from the coupled ADCIRC/WAVEWATCH III model are used for offshore boundary conditions. Several methods (2D and coupled 2D/1D setups) are examined to apply discharge along a boundary. In the coupled 2D/1D hydrodynamic model, a 1D model is applied to the rivers/tributaries and is linked to the 2D model of the nearshore domain. The performance of the D-Flow FM simulation of stream-flow and storm surge interaction is evaluated by comparing observed and simulated water levels. The model results and field observations agree well, and the measured phase and amplitude time series are well reproduced by the model. A sensitivity analysis is carried out to evaluate the effects of stream discharge and oceanic fluctuations on the model performance. The numerical results reveal that hydrodynamic predictions are dependent on the stream discharge, especially in the upstream regions. The 2D/1D hydrodynamic modeling provides sufficient dynamic information (e.g. mass and momentum) exchange between rivers and bays to account for compound flooding. These results indicate the proposed modeling framework is a good candidate for simulations of inland-coastal compound flooding under storms.