The model behaviour is primarily governed by the combination of a three basic elements:
Code, written in high tech FORTRAN90, and available from the GFDL CVS repository. The code consists of many individual modules (.F90 files) separated into ocean, atmosphere, coupler, and shared directories. The code must be compiled to make an executable. The CM2c code is identical to that used for the higher resolution GFDL IPCC-class model, CM2.
Grid specifications. These are the main difference between CM2c and CM2.
Ancillary data sets (things like runoff maps, aerosols, and seafloor roughness). Many (but not all) of these are regridded to the coarser resolution grid.
Running it requires a few additional pieces:
Initial conditions to start from
Data tables, that generally provide a few simple boundary conditions (eg. atmospheric CFC concentrations)
Diagnostic tables, that specify the output variables to be saved to disk as history files
Field tables to switch on additional ocean modules, like biogeochemistry
Optional parameter changes, provided at runtime (so that you don’t need to recompile to change them), through the namelists
How you put all the pieces together depends on whether you are running at GFDL or running elsewhere.