Repository navigation
Adjustments to ISMIP7 streams, namelist, and albany_input.yaml - #999
Merged
Merged
Conversation
trhille
changed the title
Add necessary melt parameters for ISMIP7 sub-shelf melt parameterization. Remove vestigial ismip6_melt_params stream.
Remove deprecated config_enable_timeAvgRestarts option from namelist files.
Use shorter restart intervals to reduce wasted compute time.
Fix type ("Cubrature" instead of "Cubature") in yaml file.
trhille
force-pushed
the
landice/ismip7_streams
branch
from
October 28, 2026 15:21
e0614d0 to
ac49687
Compare
* OCX forcing path * enable 3d TF * ctrl forcing paths Note that it looks like the setup function is not equipped to handle ctrl 3d TF forcing that is broken into separate files, which is the case because of large file sizes. This may be why the 3d tf had been disabled. Also note that each of the two controls must be set up separately because the cfg uses direct paths to the forcing files themselves.
matthewhoffman
force-pushed
the
landice/ismip7_streams
branch
from
October 30, 2026 18:44
7b47b80 to
47d4607
Compare
After fixing the MOLHO and FCT bugs, it became necessary to increase CD strainrate mult to 8 to match restore-calving VAF. However, at that level, Jakobhavn was exhibiting much too high and possibly borderline unstable retreat. I explored adding crevasse water depth to also increase calving. Amaral et al. (2020) recommend crevasse water depth=24 m for a GIS-wide parameter value, based on observed velocity and terminus position maps. Water depth=24m alone still greatly underestimates the VAF from restore-calving. So I then include strainrate multiplier on top of a prescribed 24 m water depth, and find strain rate multiplier of 4 works best. This commit also changes the allowable CFL fraction to 0.75 after testing shows similar accuracy to using 0.5.
matthewhoffman
changed the title
matthewhoffman
approved these changes
Nov 4, 2026
matthewhoffman
force-pushed
the
landice/ismip7_streams
branch
from
November 4, 2026 18:56
406d8d8 to
6223627
Compare
Member
|
I've noted 2 additional issues. I don't think I want to bother addressing them because they are easy to work around, but I want to flag them here.
|
That was the case in ISMIP6, but for ISMIP7, the ctrl runs are a monthly climatology, so the same stream settings apply.
The ctrl experiments were special-cased in ismip7_gris and ismip7_ais,
a holdover from the old constant-climate protocol. They used separate
ctrl_*_climatology_path options and assumed a single 3D thermal forcing
file, so the multi-year GrIS 3D TF chunks were never symlinked.
Both ctrl runs now use {forcing_basepath}/{model}_ctrl and the same
forcing-file discovery and intervals as the other experiments. In AIS,
mesh-name filtering and annual TF are kept, and ctrl still skips the
hydrofracture shelf-collapse mask. The ctrl_*_climatology_path options
are removed from the configs and the docs are updated.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
The 3D thermal forcing stream used a fixed reference_time of 2000-01-01, which puts filename boundaries at 2010, 2020, ... and does not match projection chunks that start in 2015. Set reference_time to the start of the first chunk, and fail setup if multiple files share a chunk start year (e.g. two versions of the series in one directory). Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Create graph.info.part.<ntasks> alongside graph.info for GrIS and AIS so job scripts that run MALI directly find the decomposition. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Add custom_job_script and custom_job_env_cmds options for GrIS and AIS. With custom_job_script = custom, each experiment also gets a job_script.custom.sh that runs the linked MALI executable directly. Enable it in ismip7_gris_pm_production.cfg. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
matthewhoffman
force-pushed
the
landice/ismip7_streams
branch
from
November 9, 2026 05:11
fb1a598 to
8f4f949
Compare
matthewhoffman
force-pushed
the
landice/ismip7_streams
branch
from
November 9, 2026 05:23
8f4f949 to
2e7bd7a
Compare
Another round of testing indicates this is the optimal OCX. Also set ismip7_gris_pm_production.cfg to set up OCX plus all CESM runs initially. And set account to m4274.
matthewhoffman
force-pushed
the
landice/ismip7_streams
branch
from
November 9, 2026 05:47
2e7bd7a to
14d0f7e
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This merge finalizes the streams and namelist configurations for ISMIP7.
This addresses issues tracked in #990 and #991.
The GrIS configuration in this PR is production-ready. The AIS configuration will have further adjustments in a follow-on PR.