add landice gridcomp

[weiyuan-jiang]

This PR is corresponding to the GEOSldas PR to add GEOSlake_GridComp

[gmao-rreichle]

@weiyuan-jiang: "MODEL" here is the particular flavor of the land (Catchment) model. Now that we're integrating landice and lake, which include their own "model", I think it would be good to rename "MODEL" to "LANDMODEL" for clarity. Does this make sense?

[weiyuan-jiang]

Yes. I have changed the name

[gmao-rreichle]

@weiyuan-jiang: I just came across the following assumption in pre-processing: the *.til file must have all of the land tiles at the beginning. This has worked so far, but it is not a safe assumption (as noted in my comment from 2020). You are probably already aware of it in the context of the present PR, but I'm mentioning it here just in case.

GEOSldas_GridComp/GEOSldas_App/preprocess_ldas_routines.F90

Line 2125 in 05bbcca

! assume all land tiles are at the beginning

    ! assume all land tiles are at the beginning
          ! UNSAFE ASSUMPTION! - reichle, 2 Aug 2020
          
          if (typ/=MAPL_Land .and. typ/=MAPL_Land_ExcludeFromDomain) then   ! exit if not land
             
             if (logit) then
                write (logunit,*) 'WARNING: Encountered first non-land tile in *.til file.'
                write (logunit,*) '         Stop reading *.til file under the assumption that'
                write (logunit,*) '           land tiles are first in *.til file.'
                write (logunit,*) '         This is NOT a safe assumption beyond Icarus-NLv[x] tile spaces!!'
             end if
             
             exit ! assuming land comes first in the til file
             
          end if

[gmao-rreichle]

@weiyuan-jiang: I just came across the following assumption in pre-processing...

PS: The exact same code appears again a couple of hundred lines later:

GEOSldas_GridComp/GEOSldas_App/preprocess_ldas_routines.F90

Line 2378 in 05bbcca

! assume all land tiles are at the beginning