Skip to content

Test post-processing of new h0a/h0i files (ctsm53062_f09_1850) #109

Description

@slevis-lmwg

Description:
Addresses ESCOMP/CTSM#3317

./create_newcase --case ~/cases_LMWG_dev/ctsm53062_f09_1850 --compset I1850Clm60BgcCrop -res f09_t232
to generate 20 yrs of h0a/h0i output to test our post-processing tools.

Also control simulation with ctsm5.3.061
./create_newcase --case ~/cases_LMWG_dev/ctsm53061_f09_1850 --compset I1850Clm60BgcCrop -res f09_t232
to generate 20 yrs of h0 output for one of our comparisons (expect bit-for-bit same results).


Case directory:
Locally:
~slevis/cases_LMWG_dev/ctsm53062_f09_1850
~slevis/cases_LMWG_dev/ctsm53061_f09_1850


Sandbox:
Locally:
/glade/work/slevis/git/hXi_hXa_tapes_iss1059
git branch: master
git describe: ctsm5.3.062

For the control:
/glade/work/slevis/git/b4b-dev
git branch: master
git describe: ctsm5.3.061


user_nl_ changes:
None


SourceMods:
None


Diagnostics:
Diags (if available)
https://webext.cgd.ucar.edu/I1850/$CASE/lnd/
@olyson, I think here two comparisons will be interesting:

  • "this case" against "this case" as you suggested
  • "this case" against the "control" if we find a way to make the diagnostic package backwards compatible

Output:
/glade/scratch/slevis/archive/ctsm53062_f09_1850
/glade/scratch/slevis/archive/ctsm53061_f09_1850


Contacts:
@slevis-lmwg @olyson


Extra details:

Activity

  1. slevis-lmwg commented on Jul 10, 2025

    @slevis-lmwg
    CollaboratorAuthor

    @olyson
    I ended up running ctsm5.3.062 for 20 yrs instead of 10.
    I also ended up running a control case with ctsm5.3.061.
    Output is in my /archive directory.

  2. olyson commented on Jul 10, 2025

    @olyson
    Collaborator

    ILAMB seems to run fine on ctsm53062_f09_1850 by itself with this filter specified:

    .clm2.h0a.

    However, only one filter can be specified for an ILAMB run which could be problematic if you want to compare a simulation with h0 files to a simulation with h0a files. In that case, we could use the filter:

    .clm2.h0

    (Remove the "dot" at the end). Then ILAMB should use the h0 files from, e.g., ctsm53061_f09_1850, and h0a files from, e.g., ctsm53062_f09_1850. I tested this and it seems to work fine. Output here: https://webext.cgd.ucar.edu/I1850/ctsm53062_f09_1850/lnd/_build_h0a/

    A possible problem arises if there are h0i files present AND a variable ILAMB needs is contained in both the h0a and h0i files. Currently, as noted by @slevis-lmwg , the only variable that is in the h0i file is SNOW_PERSISTENCE, which is not used by ILAMB, so not a problem. But if there is a duplicated variable we will need to resolve this, maybe by simply linking the h0a files to h0 files for that simulation.

  3. olyson commented on Jul 11, 2025

    @olyson
    Collaborator

    At @slevis-lmwg's suggestion, I've introduced "hist_ext_1" and "hist_ext_2" into the diagnostics template script. These can be set to either "h0" or "h0a". They are then used in the modules that create climatologies and trends. So this is backwards compatible.
    I've tested this on @slevis-lmwg's two simulations, ctsm53062_f09_1850 (produces h0a files), and ctsm53061_f09_1850 (produces h0 files). Seems to work fine, the two simulations are showing up as identical. Output is here:

    https://webext.cgd.ucar.edu/I1850/ctsm53062_f09_1850/lnd/ctsm53062_f09_1850_11_20-ctsm53061_f09_1850_11_20/setsIndex.html

    The new template script should be used from now on and is here:

    /glade/campaign/cgd/tss/people/oleson/FROM_LMWG/diag/lnd_diag4.2/lnd_template4.2.41.csh

    There is also a new mfiles.csh script with h0a capability added similarly:

    /glade/campaign/cgd/tss/people/oleson/FROM_LMWG/diag/lnd_diag4.2/mfiles.csh

    I still need to commit and tag this, but that requires the "Bad Gateway" issue to be fixed first.

  4. olyson commented on Jul 15, 2025

    @olyson
    Collaborator

    The "Bad Gateway" issue has been fixed, so I've committed and tagged this as lnd_diag4.2.41.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Projects

No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions