[hollyhan]

This PR adds functionality to set up coupled MALI-FastIsostasy simulations in the ISMIP7 AIS test case, following the same pattern as the existing Sea Level Model (SLM) integration.

Checklist

• User's Guide has been updated
• Developer's Guide has been updated
• API documentation in the Developer's Guide (api.rst) has any new or modified class, method and/or functions listed
• Documentation has been built locally and changes look as expected

Update (09/22/2026): Test results of coupled modeling based on realistic climate forcing and Earth sturcture can be found here: MALI-Dev/E3SM#183

July 2026: test results with a simple climate forcing

The signs of bedrock change and thickness change seem reasonable (ice thickening → bedrock subsiding & ice thinning → bedrock uplifting)

[hollyhan]

@matthewhoffman , @trhille . This is ready to be tested and merged.

[trhille]

@hollyhan, thanks for this amazing work! I have a number of comments, which I think are important but should be easy to address.

[trhille]

I tested this with MALI built at the current head of https://github.com/MALI-Dev/E3SM/tree/matthewhoffman/mali/fastisostasy (commit 1580599).

Here is my .cfg file:

Click to expand

[ismip7_run_ais]

# List of experiments to set up.
# Can be "all", "historical", "projections", "ctrl", or a
# comma-delimited list of experiment names.
# Valid experiment names:
#   historical_CESM2-WACCM, historical_MRI-ESM2-0,
#   ssp126_CESM2-WACCM, ssp126_MRI-ESM2-0,
#   ssp370_CESM2-WACCM, ssp370_MRI-ESM2-0,
#   ssp585_CESM2-WACCM, ssp585_MRI-ESM2-0,
#   ctrl_CESM2-WACCM, ctrl_MRI-ESM2-0,
#   ocx
exp_list = ocx
# historical_CESM2-WACCM,ssp585_CESM2-WACCM

# Number of tasks to use for each run
ntasks = 512

# Value to use for config_pio_stride.
# Should be divisible into ntasks
pio_stride = 128

# Base path to the pre-processed ISMIP7 forcing files.
# Expected layout:
#   {forcing_basepath}/{model}_{scenario}/atmosphere/
#   {forcing_basepath}/{model}_{scenario}/ocean_thermal_forcing/
# User has to supply.
forcing_basepath = /global/cfs/cdirs/m4288/users/trhille/ISMIP7/test_processing/AIS/

# Path to the initial condition file. User has to supply.
init_cond_path = /pscratch/sd/t/trhille/ISMIP7/AIS/mesh_gen_20260805/2008/4to20km/optimization/molho/budd/20260914/AIS_4to20km_r03_20260915.nc

# Path to the file for the basal melt parametrization coefficients.
melt_params_path = /pscratch/sd/t/trhille/ISMIP7/AIS/test_MALI-Dev_PR187_and_COMPASS_PR989/ismip6shelfMelt_params.nc

# Path to the reference surface elevation file
reference_surface_path = /pscratch/sd/t/trhille/ISMIP7/AIS/test_MALI-Dev_PR187_and_COMPASS_PR989/upper_surface.nc

# Path to the region mask file
region_mask_path = /global/cfs/cdirs/fanssie/MALI_input_files/AIS_4to20km_r03/AIS_4to20km_r03_20260905_ismip6_regionMasks.nc

# Path to the ocean thermal forcing climatology file for CTRL2015 runs.
# This is the constant-climate 30-year mean (2000-2029) thermal forcing.
# User has to supply if running ctrl experiments.
ctrl_tf_climatology_path = NotAvailable

# Path to the atmosphere climatology files for CTRL2015 runs.
# Directory containing constant-climate SMB, temperature, etc.
# User has to supply if running ctrl experiments.
ctrl_atm_climatology_path = NotAvailable

# Path to OCX (observationally constrained experiment) forcing.
# User has to supply if running the ocx experiment.
ocx_forcing_path = /global/cfs/cdirs/m4288/users/trhille/ISMIP7/test_processing/AIS/OCX_main

# Calving method to use. Options: restore, von_mises
calving_method = restore

# Path to the file containing the von Mises parameter fields.
# Only required if calving_method is set to 'von_mises'.
von_mises_parameter_path = NotAvailable

# Whether to apply the ISMIP7 Path C ice shelf collapse mask (from
# ismip7_forcing fracture) as hydrofracture-gated mask calving. Expected
# layout: {forcing_basepath}/{model}_{scenario}/shelf_collapse/*.nc
# Not used for historical, ctrl, or ocx experiments. If true and the mask
# file is missing for an experiment that should have one, setup fails.
use_hydrofracture_forcing = true


# Ice fracture toughness (K_IC, Pa m^0.5) used in the hydrofracture
# vulnerability criterion gating mask calving. Only used when
# use_hydrofracture_forcing is true.
calving_fracture_toughness = 2.0e5

gia_model = FastIsostasy

# Path to the directory containing globally defined ice thickness
# field for the sea-level model
slm_input_ice = NotAvailable

# Path to the directory containing earth model for the sea-level model
slm_input_earth = NotAvailable

# Earth structure profile filename
slm_earth_structure = prem_512.l60K2C.sum18p6.dum19p2.tz19p4.lm22

# Path to the directory containing other SLM input files
slm_input_others = NotAvailable

# Number of gauss-legendre nodes in latitude (typically multiple of 512)
nglv = 2048

fastiso_path = /global/cfs/cdirs/fanssie/MALI_projects/FastIsostasy_Inputs/

fastiso_earth_structure_filename = weak-earth.nc

fastiso_mask_filename = mask_full_domain.nc

fastiso_res_km = 16

[parallel]
parallel_executable = srun --label --cpu-bind=cores

Bed topography change 2008–2012 (not 100% sure which way the sign points because of the way Paraview programmable filter works):

[trhille]

Note that we're getting a little ahead of ourselves by merging this before merging the MALI-Dev PR MALI-Dev/E3SM#183. I should have checked that this works with gia_model = none before merging this. I just tested that, and everything seems to work as expected.