Hydrodynamic modeling of a complex salt marsh system: Importance of channel shoreline and bathymetric resolution

Abstract

Modeling hydrodynamics and sediment transport inside a multi-inlet wetland system is a challenging task due to constraints on model efficiency, accuracy and representation of physics, the scarcity of field data for model validation, and more importantly, the availability of high-resolution data sets of channel bathymetry. Lack of field data sets, model assumptions, and limitations often lead to wrong interpretation of the system’s nature. To correctly predict the changes in governing hydrodynamics and morphology of shallow wetland environments, it is essential to resolve the channel properties with the best possible accuracy. In this study, we consider the importance of two factors, numerical model grid resolution around channel shorelines, and bathymetry surveys, in improving the interior channel and mudflat hydrodynamics. Channel shoreline element size is observed to be an essential factor in time-varying channel volume estimation. Lower model resolution in channels inside a marsh can provide good agreement with in-situ surface elevation data, even though the channel surface-velocity phase lag, and velocity magnitude and asymmetry are inaccurate. A higher grid resolution at the channel berms and the high-density channel survey help define the time-varying lateral shoaling bathymetry, cross-section area, and bottom friction, and improve the corresponding model skill and tidal wave properties significantly. We have selected Bombay Hook National Wildlife Refuge, a rapidly eroding salt-marsh system in Delaware Bay, to compare model skills for different hypsometric conditions. Model performance is evaluated during two storm conditions; Hurricane Sandy (2012) and Hurricane Joaquin (2015), using tide gauges located on the marsh interior channels. • Poor grid resolution around marsh-channel berm creates artificial channel convergence and inaccurate tidal wave characteristics. • Topo-bathymetry field survey and bias correction improves channel hypsometry and numerical model reliability significantly. • Inaccurate channel bathymetry and inadequate resolution might lead to acceptable surface elevation but poor tidal velocity field. • The conventional grid generation methods based on sparse cross channel surveys is not sufficient for wetland models.