List of netCDF file attributes that are incompatible with MAPL

[yantosca]

This PR is intended to document the various netCDF file attributes that cause MAPL to hang (see issue geoschem/GCHP#373).

1. Any field that is of integer type, e.g. time(int)

2. bnds, lat_bnds, lon_bnds, lev_bnds. However, I wonder if this is more because lon_bnds, lat_bnds, lev_bnds do not have long_name and units attributes. I can test this.

   dimensions:
       time = UNLIMITED ; // (1 currently)
       lon = 72 ;
       bnds = 2 ;  <=== incompatible with GCHP
       lat = 46 ;
       lev = 72 ;
   
       double lon_bnds(lon, bnds) ;
               lon_bnds:_Storage = "chunked" ;
               lon_bnds:_ChunkSizes = 72, 2 ;
               lon_bnds:_DeflateLevel = 1 ;
               lon_bnds:_Endianness = "little" ;
               lon_bnds:_NoFill = "true" ;
       double lat_bnds(lat, bnds) ;
               lat_bnds:_Storage = "chunked" ;
               lat_bnds:_ChunkSizes = 46, 2 ;
               lat_bnds:_DeflateLevel = 1 ;
               lat_bnds:_Endianness = "little" ;
               lat_bnds:_NoFill = "true" ;
       double lev_bnds(lev, bnds) ;
               lev_bnds:standard_name = "atmosphere_hybrid_sigma_pressure_coordinate" ;
               lev_bnds:units = "level" ;
               lev_bnds:formula_terms = "ap: ap_bnds b: b_bnds ps: ps" ;
               lev_bnds:_Storage = "chunked" ;
               lev_bnds:_ChunkSizes = 72, 2 ;
               lev_bnds:_DeflateLevel = 1 ;
               lev_bnds:_Endianness = "little" ;
               lev_bnds:_NoFill = "true" ;

3. Area variable without a time dimension:

       float AREA(lat, lon) ;
               AREA:long_name = "Surface area" ;
               AREA:units = "m2" ;
               AREA:_Storage = "chunked" ;
               AREA:_ChunkSizes = 46, 72 ;
               AREA:_DeflateLevel = 1 ;
               AREA:_Endianness = "little" ;
               AREA:_NoFill = "true" ;

4. Any variable that does not have the same number of vertical levels as the lev dimension.

More to follow

Tagging @lizziel

[lizziel]

Transferring to GCHP so we can keep GCHP issues all in the same place.

[lizziel]

I think #1 needs to be checked again since @msulprizio pointed out the input meteorology files have time as integer without issue.

[lizziel]

GMAO is putting together a guide on netCDF file requirements for MAPL: https://github.com/GEOS-ESM/MAPL/wiki/Guide-to-Supported-ExtData-Input-Files

[lizziel]

ReadTheDocs Debugging page now links to the GMAO wiki page for MAPL-supported input files.