generate square grid SCRIP files for coupling FastIsostasy to MALI

[hollyhan]

Add Python script to create square grid SCRIP files for
FastIsostasy MALI coupling

Summary

• Add create_square_grid_scripfile.py tool to generate
  SCRIP-format files for FastIsostasy coupling
• FastIsostasy requires a perfect square grid (nx × nx)
  for bedrock deformation coupling
• Script reads MALI mesh SCRIP file, extracts domain
  bounds, and creates a square regular lat-lon grid
• Enables users to generate mapping weight files with
  ESMF_RegridWeightGen

Motivation

FastIsostasy coupling in MALI requires two mapping weight
files:

• mpas_to_fastisostasy.nc (MALI mesh → square grid)
• fastisostasy_to_mpas.nc (square grid → MALI mesh)

The target grid must be a perfect square (code checks `nX

• nX = nGridinmpas_li_bedtopo.F:1021-1027`). This
  script automates the creation of the required square grid
  SCRIP file.

Changes

• New file: create_square_grid_scripfile.py
  • Reads MALI SCRIP file to determine domain extent
  • Creates nx × nx square grid with proper SCRIP format
  • Supports optional padding around domain
  • Includes usage documentation and example workflow

Usage Example

# Create 512×512 square grid for FastIsostasy
python create_square_grid_scripfile.py \
    --mali_scripfile mali_mesh.scripfile.nc \
    --nx 512 \
    --output fastisostasy_512x512.scripfile.nc

# Generate mapping weights with ESMF
ESMF_RegridWeightGen -s mali_mesh.scripfile.nc -d
fastisostasy_512x512.scripfile.nc \
    -w mpas_to_fastisostasy.nc -i --64bit_offset
--src_regional


Grid comparison:

┌─────┬──────────────┬──────────┬─────────────────────┐
│ nx  │     Lat      │  Memory  │      Use case       │
│     │  resolution  │          │                     │
├─────┼──────────────┼──────────┼─────────────────────┤
│ 128 │ ~24 km       │ Small    │ Quick testing       │
├─────┼──────────────┼──────────┼─────────────────────┤
│ 150 │ ~20 km       │ Small    │ Testing             │
├─────┼──────────────┼──────────┼─────────────────────┤
│ 256 │ ~12 km       │ Medium   │ Moderate runs       │
├─────┼──────────────┼──────────┼─────────────────────┤
│ 384 │ ~8 km        │ Large    │ Production          │
│     │              │          │ (stability limit)   │
├─────┼──────────────┼──────────┼─────────────────────┤
│ 512 │ ~6 km        │ Very     │ Too fine (unstable) │
│     │              │ large    │                     │
└─────┴──────────────┴──────────┴─────────────────────┘

Test Plan

- [ ] Run script on example MALI mesh SCRIP file
- [ ] Verify output SCRIP file has correct dimensions (nx
= ny)
- [ ] Generate weight files with ESMF_RegridWeightGen
- [ ] Test weight files in MALI run with
config_uplift_method='fastisostasy'

🤖 Generated with Claude Code

[hollyhan]

Hi @matthewhoffman,

This PR ready to be reviewed.

A sample test run result is shown below:

(e3sm-unified) hollyhan@login40:/pscratch/sd/h/hollyhan> python ~/scripfile_for_fastisostasy/create_square_grid_scripfile.py --mali_scripfile ~/scripfile_for_fastisostasy/relaxed_10yrs_4km.nc.scripfile --nx 384 --output fastisostasy_384x384.scripfile.nc
Reading MALI scripfile: /global/homes/h/hollyhan/scripfile_for_fastisostasy/relaxed_10yrs_4km.nc.scripfile
  Domain: lat [-89.93, -62.82]°
          lon [0.00, 360.00]°
  Longitude convention: [0, 360]
Creating 384x384 square grid:
  Resolution: dlat=0.0708°, dlon=0.9399°
  Total cells: 147456
Writing SCRIP file: fastisostasy_384x384.scripfile.nc
SUCCESS: Created fastisostasy_384x384.scripfile.nc

Next steps:
1. Generate weight files with ESMF_RegridWeightGen:
   # MALI to FastIsostasy
   ESMF_RegridWeightGen -s /global/homes/h/hollyhan/scripfile_for_fastisostasy/relaxed_10yrs_4km.nc.scripfile -d fastisostasy_384x384.scripfile.nc \
       -w mpas_to_fastisostasy.nc -i --64bit_offset --src_regional --dst_regional
   # FastIsostasy to MALI
   ESMF_RegridWeightGen -s fastisostasy_384x384.scripfile.nc -d /global/homes/h/hollyhan/scripfile_for_fastisostasy/relaxed_10yrs_4km.nc.scripfile \
       -w fastisostasy_to_mpas.nc -i --64bit_offset --src_regional --dst_regional

[matthewhoffman]

@hollyhan , thanks for getting this rolling! I have a couple thoughts to run by you. Could we instead use the projected ISMIP diagnostic grid at 8km (or any resolution) rather than the custom regular grid produced by this PR? It seems like the approach in this PR will result in a somewhat inexact grid spacing, and it might be useful to have the GIA results on a standardized grid. If we did that, we could use the existing scrip file creation tool (https://github.com/MPAS-Dev/MPAS-Tools/blob/master/conda_package/mpas_tools/scrip/from_planar.py) and not need a specialized one just for this MALI application. What do you think?

@xylar , do you you have input on this?

[xylar]

Seems fine to me at a quick glance. Let me know if you want more specific feedback.

[xylar]

do you you have input on this?

Maybe you meant feedback on your comment to Holly. If so, I agree that this tool might be redundant with ones we already have. It might be better to see if the existing tools work. And I agree that using a standard grid (e.g. ISMIP) might be best.

[hollyhan]

Thanks so much, @matthewhoffman and @xylar for your comments!

Somehow it didn't appear to me I could have used the existing tool! I totally agree with you it might be best to use the ISMIP standardized grid. I'll close this PR upon testing the coupled MALI-FI simulation with based on the scrip file generated using the from_planar.py script.

[matthewhoffman]

@hollyhan , that sounds good. Maybe it would make sense to put the FI set up entirely in the ismip7_run test case - there could be an option added there to use FI, and if true, it would set up the FI grid to match one of the standard ISMIP7 regular grids, and then generate the scrip and mapping files (similar to how the ismip6_run test case set up the SLM).

You could copy this function to the ismip7_run test case for setting up the regular grid:
https://github.com/MPAS-Dev/MPAS-Tools/pull/733/changes#diff-ad34211ba20a8ac13dfce99f110e5a7349da732d33ffb81b06aa542455ba541eR147