Wave Model

In order to explore the Long Island Sound (LIS) circulation and the wave dynamics at the coastal area, ran a numerical hydrodynamic- wave coupled model in the region.  The model corresponds to the Finite Volume Community Ocean Model (FVCOM) (Chen et al., 2006a; Chen et al., 2006b; Chen et al., 2007).  The model uses a 3-D unstructured grid, free-surface and primitive equations to calculate the hydrodynamics.  The FVCOM relies on the wet/dry point treatment method to calculate water transport into and out of the intertidal zone.  This circulation model was coupled with the wave module SWAVE, which was developed by Qi et.al (2009) and corresponds to an unstructured finite version of the SWAN (Simulating Waves Nearshore) model developed by the SWAN team at the Delft University of Technology.  The SWAN model (which follows from the HISWA wave model) was developed specifically to resolve the smaller coastal scales, where shallow water processes are in effect.  The hydrodynamics and waves are coupled via radiation stresses, bottom boundary layer and surface stress (e.g., Wu et al., 2011).  To our knowledge this is the first effort to implement and validate a fully coupled high-resolution hydrodynamic- wave coupled model in LIS.

The model was forced at the open boundary with the harmonic tidal components (Foreman 1978).  Riverine discharge was limited to the Connecticut River.  Atmospheric forcing follows from North American Mesoscale Forecast System (NAM) and Weather Research and Forecasting (WRF) model simulations, where at this point only wind vector fields are included (i.e. no heat fluxes and no precipitation).  The spin up for the model ranges between 1 – 2 days.  The horizontal grid resolution (Dx, Dy) is 250 m with 11 sigma layers in the vertical (z).  The turbulent closure model corresponds to the q-ql Mellor and Yamada (1982) level 2.5 (i.e. MY-2.5), where q is the turbulent kinetic energy and l is the turbulent scale.

The bottom boundary layer follows a law-of-the-wall classic behavior, where the bottom drag coefficient follows by matching a logarithmic profile at height zab (depending on model resolution) and the the aerodynamic roughness length (zo) :


where k correspond to the von Karman constant (k = 0.4).