Update orography code for low-res grids

[GeorgeGayno-NOAA]

DESCRIPTION OF CHANGES:

Low-resolution grids are defined as C24/C18/C12.

1 - Updates to the orography code:

• read_global_orog and read_global_mask - A check was added to ensure the i/j-dimensions of the global data match the dimensions passed in as arguments.
• qc_orog_by_ramp was updated to read the j-dimension of the RAMP data from the file itself, instead of being declared in the routine. This made writing a unit test easier.
• makemt2 - variable "maxsum" is now calculated dynamically instead of being hardwired. The hardwired value was not large enough for grids coarser than C48.
• make_mask - variable "maxsum" is not used and was removed.
• remove_isolated_pts - This routine was originally written for the global gaussian grid. Some of the logic did not work for FV3 tiles. To fix, the check along the "east/west" tile edges was removed. Also removed was logic that was (I believe) used for reduced gaussian grids (where the number of 'i' points decreased toward the poles).
• All I/O routines were updated to use the F90 NetCDF routines.
• Minor corrections to the Doxygen for a few routines.
• The gdas_init utility was updated to set the "fixed" file directory for C24/C18/C12.

2 - New unit tests for the following routines/functions:

• find_nearest_pole_points
• find_poles
• get_index
• get_lat_angle
• get_lon_angle
• get_xnsum
• get_xnsum2
• get_xnsum3
• inside_a_polygon
• latlon2xyz
• minmax
• qc_orog_by_ramp
• read_global_mask
• read_global_orog
• read_mask
• read_mdl_dims
• read_mdl_grid_file
• remove_isolated_pts
• transpose_mask
• transpose_orog

TESTS CONDUCTED:

• Compile branch on all Tier 1 machines using Intel (Orion, Jet, Hera, Hercules and WCOSS2). Done using 64d26a4.
• Compile branch on Hera using GNU. Done using 64d26a4.
• Compile branch in 'Debug' mode on WCOSS2. Done on Cactus using 64d26a4. There were no warnings.
• Run unit tests locally on any Tier 1 machine using both Intel and GNU. Done on Cactus (Intel) and Hera (GNU) using 64d26a4. All tests passed.
• Run grid_gen consistency tests locally on all Tier 1 machines. Done on Hera, Hercules, Jet, Cactus and Orion using 64d26a4. All tests passed as expected.
• Compile with Doxygen. Done on Cactus using 64d26a4. The orog.fd code had no warnings.

Describe any additional tests performed.

• Using 64d26a4, C12/18/24 uniform test grids were created on Hera - see /scratch1/NCEPDEV/da/George.Gayno/ufs_utils.git/UFS_UTILS/fix/orog.lowres
• Using 64d26a4 on Hera, gdas_init utility was tested on a C12 grid. See: Update orography code for low-res grids #1012 (comment)

DEPENDENCIES:

None.

DOCUMENTATION:

Doxygen successfully created. See above.

ISSUE

Fixes #1000.

RELATED ISSUES

ufs-community/ufs-weather-model#2508

[GeorgeGayno-NOAA]

The gdas_init utility was run on Hera for a 2025/01/08/06z C12 case. The 'fixed' files for C12 were staged in ./orog.lowres here:

/scratch1/NCEPDEV/da/George.Gayno/ufs_utils.git/UFS_UTILS/fix $ ls -l
total 8
lrwxrwxrwx 1 George.Gayno da   48 Sep 30 13:49 am -> /scratch1/NCEPDEV/global/glopara/fix/am/20220805
-rwxr-xr-x 1 George.Gayno da 2247 Nov 19 16:48 link_fixdirs.sh
lrwxrwxrwx 1 George.Gayno da   11 Jan 15 14:44 orog -> orog.lowres
drwxr-sr-x 5 George.Gayno da 4096 Nov 21 19:14 orog.lowres
lrwxrwxrwx 1 George.Gayno da   55 Sep 30 13:49 sfc_climo -> /scratch1/NCEPDEV/global/glopara/fix/sfc_climo/20230925

/scratch1/NCEPDEV/da/George.Gayno/ufs_utils.git/UFS_UTILS/fix/orog.lowres/C12 $ ls -l
total 836
-rw-r--r-- 1 George.Gayno da 40615 Nov 22 17:48 C12.mx100_oro_data.tile1.nc
-rw-r--r-- 1 George.Gayno da 40615 Nov 22 17:48 C12.mx100_oro_data.tile2.nc
-rw-r--r-- 1 George.Gayno da 40615 Nov 22 17:48 C12.mx100_oro_data.tile3.nc
-rw-r--r-- 1 George.Gayno da 40615 Nov 22 17:48 C12.mx100_oro_data.tile4.nc
-rw-r--r-- 1 George.Gayno da 40615 Nov 22 17:48 C12.mx100_oro_data.tile5.nc
-rw-r--r-- 1 George.Gayno da 40615 Nov 22 17:48 C12.mx100_oro_data.tile6.nc
-rw-r--r-- 1 George.Gayno da 40615 Dec 10 19:28 C12.mx900_oro_data.tile1.nc
-rw-r--r-- 1 George.Gayno da 40615 Dec 10 19:28 C12.mx900_oro_data.tile2.nc
-rw-r--r-- 1 George.Gayno da 40615 Dec 10 19:28 C12.mx900_oro_data.tile3.nc
-rw-r--r-- 1 George.Gayno da 40615 Dec 10 19:28 C12.mx900_oro_data.tile4.nc
-rw-r--r-- 1 George.Gayno da 40615 Dec 10 19:28 C12.mx900_oro_data.tile5.nc
-rw-r--r-- 1 George.Gayno da 40615 Dec 10 19:28 C12.mx900_oro_data.tile6.nc
-rw-r--r-- 1 George.Gayno da 51123 Nov 21 19:14 C12_grid.tile1.nc
-rw-r--r-- 1 George.Gayno da 51123 Nov 21 19:14 C12_grid.tile2.nc
-rw-r--r-- 1 George.Gayno da 51123 Nov 21 19:14 C12_grid.tile3.nc
-rw-r--r-- 1 George.Gayno da 51123 Nov 21 19:14 C12_grid.tile4.nc
-rw-r--r-- 1 George.Gayno da 51123 Nov 21 19:14 C12_grid.tile5.nc
-rw-r--r-- 1 George.Gayno da 51123 Nov 21 19:14 C12_grid.tile6.nc
-rw-r--r-- 1 George.Gayno da 22865 Nov 21 19:14 C12_mosaic.nc
drwxr-sr-x 7 George.Gayno da  4096 Dec  9 19:40 ocean_mask
-rw-r--r-- 1 George.Gayno da   317 Dec 10 19:29 readme.C12.mx900.txt
drwxr-sr-x 2 George.Gayno da 12288 Dec 10 19:28 sfc

The utility ran to completion and successfully created model coldstart files.

[sanatcumar]

Hi George,
we will need the C12/18 and C24 ocean mask to test this change fully?? or we can test just the C96 to show it works ?. At some point should I generate the C12/18 an 24 ocn mask files ?

[GeorgeGayno-NOAA]

Hi George, we will need the C12/18 and C24 ocean mask to test this change fully?? or we can test just the C96 to show it works ?. At some point should I generate the C12/18 an 24 ocn mask files ?

Denise created ocean mask files for those resolutions. And I created grids using those mask files. Everything is on Hera: /scratch1/NCEPDEV/da/George.Gayno/ufs_utils.git/UFS_UTILS/fix/orog.lowres. Her code changes will be added in a subsequent PR.

These grids are being created for low resolution UFS weather model tests. I don't know if they will be officially baselined in the 'fix' directory.

This PR fixes a problem in the orog code that caused a crash at these resolutions. While fixing that error, I added some additional unit testing.

[sanatcumar]

Tested the latest commit. I was able to compile with no errors. tested a default C96.mx100 grid. All grid generation related functions are unaffected. The gdas_init utility ran successfully and created model cold start files as expected for C12 and C24 cases.