produce tile nc4 file while skipping land params generation

[weiyuan-jiang]

No description provided.

[weiyuan-jiang]

This may have duplicated the computation

[weiyuan-jiang]

There is nonzero diff for elevation variable. The orignal Nc4 file get raster_id from the memory but the TileFile_ASCII_to_Nc4 reads it from raster file, which leads to nonzero diff

[biljanaorescanin]

For me for v14 for these resolutions
CF0180x6C_DE1440xPE0720
CF0180x6C_M6TP1440x1080
CF0540x6C-SG001_CF0540x6C
DC0576xPC0361_DE0360xPE0180
EASEv2_M09
just tile option, so all under geometry, is zero diff.

I also tried option of full run and that works as before.

@weiyuan-jiang did you try to run NL3 and got differences?
Those are different due to compiler when they were created. They are quite old. We can't reproduce those. We even have note about that in the package.

GEOSgcm_GridComp/GEOSagcm_GridComp/GEOSphysics_GridComp/GEOSsurface_GridComp/Utils/Raster/makebcs/make_bcs_questionary.py

Lines 182 to 191 in 899cb83

	When produced by this code, BCs labeled with 'archived*' below will differ from BCs in archived directories!!! \n \
	These differences are caused by compiler changes, code improvements and bug \n \
	fixes that were implemented since the archived BCs in the above-mentioned \n \
	directories were originally created. The impact of these differences on \n \
	science is insignificant, and the parameter files produced by current \n \
	code are scientifically equivalent to corresponding archived BCs. \n \
	Current code always applies \"mean land elevation fix\" of v13. For v12 and earlier, land elevation produced \n \
	by current code thus differs from that in bcs_shared project disk. \n",
	"choices": [ \
	"NL3 : Icarus-NLv3 (archived*: /discover/nobackup/projects/gmao/bcs_shared/legacy_bcs/Icarus-NLv3/)", \

[weiyuan-jiang]

@biljanaorescanin @gmao-rreichle I have moved the lake_type to MAPL. So the PR needs this one from MAPL . Also, please note that I changed -9999 to MAPL_UNDEFINED_INTEGER to make it consistent with other integer types. Would this change be acceptable ? @lcandre2