Migrate initial condition adjustments/modifications from the dycore to chgres_cube

[JeffBeck-NOAA]

Problem: FV3 currently makes scientific assumptions about IC and LBC data based solely on the content of the global attribute “source” of the netCDF files from chgres_cube. This makes it impossible to address these various differences among input data sources (e.g., GFS nemsio vs. GFS grib2 vs. HRRR grib2) with sufficient granularity.

Solution: Move any and all data reformulations from the FV3 dycore (external_ic.F90 & fv_regional_bc.F90) into the appropriate locations in chgres_cube where more information about the input source is available.

See planning document for this change here.

[LarissaReames-NOAA]

Note that solving this Issue will also involve coordinated changes to the FV3 code base itself.