The model is free, with source code made available by request via the website,[1] allowing users to run the model on any system with a
Fortran compiler. A pre-compiled
Windows version of the model can also be purchased alongside the
SMS software.[8] ADCIRC is coded in
Fortran, and can be used with native
binary,
text, or
netCDF file formats.
Capabilities
The model formulation[9]
is based on the
shallow water equations, solving the
continuity equation (represented in the form of the Generalized Wave Continuity Equation[10])
and the momentum equations (with
advective,
Coriolis,
eddy viscosity, and
surface stress terms included). ADCIRC utilizes the
finite element method in either three-dimensional or two-dimensional depth-integrated form on a triangular
unstructured grid with
Cartesian or
spherical coordinates. It can run in either
barotropic or
baroclinic modes, allowing inclusion of changes in water density and properties such as salinity and temperature. ADCIRC can be run either in serial mode (e.g. on a personal computer) or in parallel on
supercomputers via
MPI. The model has been optimized to be
highly parallelized, in order to facilitate rapid computation of large, complex problems.[11][12]
ADCIRC is able to apply several different bottom friction formulations including
Manning's n-based bottom drag due to changes in land coverage (such as forests, cities, and seafloor composition), as well as utilize atmospheric forcing data (wind stress and atmospheric pressure) from several sources, and further reduce the strength of the wind forcing due to
surface roughness effects.[13][14]
The model is also able to incorporate effects such as time-varying topography and bathymetry, boundary fluxes from rivers or other sources, tidal potential, and sub-grid scale features like levees.
ADCIRC is frequently coupled to a
wind wave model such as STWAVE,
SWAN, or
WAVEWATCH III, especially in storm surge applications where
wave radiation stress can have important effects on ocean circulation and vice versa. In these applications, the model is able to take advantage of tight coupling with wave models to increase calculation accuracy.[14][15]